Download This PDF is a selection from a published volume from... National Bureau of Economic Research

This PDF is a selection from a published volume from the
National Bureau of Economic Research
Volume Title: Monetary Policy under Very Low Inflation
in the Pacific Rim, NBER-EASE, Volume 15
Volume Author/Editor: Takatoshi Ito and Andrew K. Rose,
editors
Volume Publisher: University of Chicago Press
Volume ISBN: 0-226-37897-7
Volume URL: http://www.nber.org/books/ito_06-1
Conference Date: June 25-27, 2004
Publication Date: September 2006
Title: Two Decades of Japanese Monetary Policy and the
Deflation Problem
Author: Takatoshi Ito, Frederic S. Mishkin
URL: http://www.nber.org/chapters/c0092
4
Two Decades of Japanese Monetary
Policy and the Deflation Problem
Takatoshi Ito and Frederic S. Mishkin
4.1 Introduction
The Japanese economy has been stagnant for more than ten years. The
average growth rate from 1993 to 2003 is just above 1 percent. Since 1998,
the inflation rate, either measured by gross domestic product (GDP) deflator or Consumer Price Index (CPI), has been negative. The deflation has
brought the CPI price level by the end of 2003 to 3 percent below the 1997
level. During the same period, the U.S. CPI has increased by 12 percent.
Due to virtually zero growth and deflation, the Japanese nominal GDP had
shrunk by 4 percent from 1997 to 2002, while during the same period, nominal GDP of the United States has increased by 25 percent.
Many problems have been pointed to as contributing factors that explain the “lost decade” in Japan. The burst bubble and the nonperforming
loans problem are often blamed for the poor performance of the early stage
of the stagnation. By 2003, land and stock price indexes have fallen to
between one-third and one-fourth of the respective peak in 1989–91. Slow
policy responses to the nonperforming loan problem resulted in the banking crisis of 1997–98, and the financial sector is still weak. The consumpTakatoshi Ito is a professor in the Research Center for Advanced Science and Technology
at the University of Tokyo, and a research associate of the National Bureau of Economic Research. Frederic S. Mishkin is the Alfred Lerner Professor of Banking and Financial Institutions at the Graduate School of Business, Columbia University, and a research associate of
the National Bureau of Economic Research.
This paper was written for the NBER 15th East Asian Seminar on Economics, June 25–27,
2004. The authors are grateful to Takeshi Kudo and Emilia Simeonova for their excellent research assistance. We also thank our discussants Ken Kuttner, Kazuo Ueda, Kunio Okina,
and participants at seminars at the Bank of Japan, and the East Asian Seminar on Economics. Any views expressed in this chapter are the views of the authors only and not the University of Tokyo, Columbia University, or the National Bureau of Economic Research.
131
132
Takatoshi Ito and Frederic S. Mishkin
tion tax rate increase and repeal of income tax cut in April 1997 is often regarded as a fiscal policy mistake. Slow structural reform in regulated sectors is another problem for the Japanese economy, which has not benefited
from the information and communication technology (ICT) advances that
propelled the U.S. economy.
But the most likely cause for deflation in Japan is a failure of monetary
policy, since inflation or deflation is ultimately a monetary phenomenon.
The Bank of Japan (BOJ) was unable to stop the inflation rate from turning negative, despite its various efforts. The uncollateralized call rate (the
policy interest rate that corresponds to the Federal Funds rate of the
United States) was lowered to virtually zero in February–March 1999. The
BOJ raised the call rate to 0.25 percent in August 2000 in false expectation
of continuing economic expansion, against protests from the government
and many economists. The interest rate was lowered to zero again in March
2001, with an additional measure of quantitative easing, setting the target
of current account (reserves) of commercial banks at the BOJ in excess of
required reserves. The target amount of current account was initially set at
5 trillion yen, while the required reserves was about 4 trillion yen. The target amount has been raised in several steps to a range of 30–35 trillion yen
by January 2004. In addition to raising the target amount of current account at the BOJ, the bank expanded the amount of monthly outright purchase of long-term government bonds from 400 billion yen to 600 billion
yen in August 2001, and in several steps to 1,200 billion yen in October 2002.
In addition, purchases of some of private debts, including asset-backed securities (ABS), have been introduced.
By 2002, the economy and the financial institutions weakened again.
Deflationary expectations were setting in, and consumption and investment were depressed. Aggregate demand fell short of potential output, and
the widened output gap depressed prices, reinforcing deflationary expectations. There did not seem to be a solution to the deflationary spiral.
When the zero interest rate policy (ZIRP) was first introduced in February 1999, it was intended to continue until “deflationary concern is dispelled.” It was then lifted in August 2000. When it was reintroduced in
March 2001, it was declared to continue until “the inflation rate becomes
stably above zero.” The condition was further elaborated in October 2003,
so that the necessary condition for the exit from ZIRP is that the CPI inflation rate becomes zero or above for a few months and there was no forecast by the board members of falling back to deflation. The determination
to fight deflation seems to have been strengthened. Given that deflation
was not over at the time of ZIRP termination and that the ZIRP had to be
reinstated, the interest rate hike of August 2000 was clearly a mistake.
Lively debates have taken place as to what the BOJ could have done to
prevent deflation from occurring and getting worse, and on what the BOJ
could do to get out of deflation. Many academics and policymakers, in-
Two Decades of Japanese Monetary Policy and the Deflation Problem
133
cluding studies at the Federal Reserve Board, argue that the Bank of
Japan’s actions were too little too late, at least in retrospect, in preventing
deflation from emerging and fighting out of deflation. Many academic critics have been arguing for nonconventional monetary policy in combating
deflation: for example, purchasing long-term bonds, equities, foreigncurrency-denominated bonds, and nonperforming loans. However, it has
been pointed out that the transmission channel of nonconventional monetary policy is unclear.
Inflation targeting has also been proposed as a tool to promote an independent central bank and to help get out of deflation. Namely, a credible
announcement of inflation targeting, of say 1 to 3 percent, would make inflation expectations become higher, so that the deflationary spiral would be
broken. A combination of inflation targeting as a communication and anchoring device with nonconventional policies was advocated by academic
work in the past.1
However, the BOJ has opposed inflation targeting, with economists in
the Bank of Japan arguing that there are no clear instruments to get out of
deflation, and a mere announcement without instruments would not convince market participants to change their inflation expectations. But, others in the bank have suggested that the commitment to keep the zero interest rate policy until the inflation rate becomes stably above zero has similar
effects to inflation targeting.
The chapter is organized as follows. Sections 4.2 and 4.3 will review Japanese monetary policy over the last two decades. The former concentrates
on the period of bubble and burst (1985–1997), and the latter examines the
issue under the new law of the BOJ (1998–). Section 4.4 discusses whether
estimates of Taylor rules can be used to assess Japanese monetary policy.
Section 4.5 discusses the costs of deflation. Section 4.6 examines monetarypolicy actions to prevent deflation, and Section 4.7 surveys the literature on
monetary policy to cure deflation and discusses nonconventional monetarypolicy measures. Section 4.8 concludes the chapter.
4.2 Monetary Policy and the Bubble
4.2.1 Bubble and Burst
Some researchers go back to the bubble period, 1985–90, as a source of
the Japanese stagnation in the 1990s. Since the bubble occurred and burst,
the Japanese economy fell into a difficult position of having nonperforming loans that led to the banking crisis. Some economists seem to believe
that there was a mistake in monetary policy in the 1980s, and once the burst
1. See Ito (2000, 2001), Kazumasa Iwata (2002), and Kikuo Iwata (2001, 2002) in the books
written in Japanese.
134
Fig. 4.1
Takatoshi Ito and Frederic S. Mishkin
Land prices and stock prices, 1970–2003
bubble occurred, monetary policy became powerless in the 1990s, because
the transmission channel from the interest rate policy to the real economy
was no longer operational. Therefore, it is entirely appropriate to start the
story of deflation from the bubble years.
In retrospect, it is obvious that the Japanese economy was experiencing
a bubble economy: the stock price index and the land price index quadrupled from 1983 to 1989. The stock prices index (Nikkei 225) rose from
10,000 yen at the end of 1983 to near 40,000 at the end of 1989. The economic growth rate was approaching 5 percent, surpassing the average of 4
percent from 1975 to 1989, and the tax revenues were increasing to close a
fiscal gap that had plagued the economy for two decades. At the end of the
1980s, many economists as well as policymakers around the world were
praising the Japanese economy for its excellent performance.2 Although a
few economists raised concerns, many financial analysts and bankers were
not alarmed at the apparent high value of stocks and land compared to
their cash-flow earning. Land and stock price movements from 1970 to
2003 are shown in figure 4.1.
The inflation rate had gradually come down from 12 percent in 1974 to
below 4 percent in 1978. The inflation rate suddenly went up to about 8 percent in 1979 due to the second oil crisis. However, the CPI inflation rate was
quickly brought down to below 3 percent in 1982. The inflation rate fluctuated at the low range of 0–3 percent for the rest of the 1980s. The infla2. See Ito (1992) for a comprehensive explanation of the Japanese economy up to 1991.
Two Decades of Japanese Monetary Policy and the Deflation Problem
Fig. 4.2
135
CPI inflation excluding fresh food and consumption tax, 1985–2002
tion performance of Japan from 1976 to 1987, despite a lack of legal independence of the BOJ, was often praised in the literature.3 Figure 4.2 shows
the CPI inflation rate (excluding fresh food), that is adjusted for the consumption tax introduction in 1989, and the consumption tax rate increase
in 1997.
It is remarkable that during the bubble period, the CPI inflation rate remained low. While asset prices were doubling and tripling in a few years,
the CPI inflation rate remained quite reasonable, prompting a difficult
choice to the BOJ. Indeed, the BOJ did not start tightening until 1989. Although the BOJ would not target asset prices, the burst bubble would make
monetary policy more difficult—all with the benefit of hindsight. The yen
appreciated from 260 yen/dollar in February 1985 to 150 yen/dollar in the
summer of 1986, of which some part was a movement toward an equilibrium and some part was overshooting. The sharp yen appreciation caused
a recession (due to a slump of exports) and imported disinflation. Interest
rates were lowered from 1986 to 1987 in part to help stimulate the economy
that was depressed by sharp yen appreciation.4 Low interest rates were necessary to prevent the yen from appreciating too much.
Monetary policy was finally tightened in 1989. The official discount rate
(ODR) rose from 2.5 percent, where it had been since 1987, to 3.25 percent
in May 1989. The ODR rose to 3.75 percent in October and 4.25 percent in
December. Despite this rapid hike of the interest rate, the CPI inflation rate
3. See, for example, Cargill, Hutchison, and Ito (1997), for the view that the BOJ might have
had de facto independence and exercised it wisely.
4. See, for example, Okina, Shirakawa, and Shiratsuka (2001) for such a view. They seem to
blame international policy coordination, such as the Plaza Accord of September 1995 and the
Louvre Accord of February 1997, for the BOJ not acting in a timely manner.
136
Takatoshi Ito and Frederic S. Mishkin
rose from 1 percent at the beginning of 1989 to 3 percent toward the end of
the same year. The official discount rate was raised to 6.00 percent in August 1990 (a 350 basis point hike in fifteen months).
Stock prices peaked at the level of 39,000 in Nikkei 225 index at the end
of 1989. In tandem with the interest rate hike, regulatory tightening was
applied to stop increases in land prices including: limiting the increase in
bank lending to real estate related projects and companies in the spring of
1990, and raising taxes on realized capital gains from land investment.
Stock prices finally turned down from the first trading day of 1990. The
stock price index declined by one-third from the end of 1989, the peak, to
the end of 1990. Stock prices continued to decline and the index lost 60 percent of the peak level by the summer of 1992. Land prices started to decline
in 1991. The bubble had burst.
One may question whether monetary policy was too lax for too long
during the bubble experience, that is the second half of the 1980s. What if
monetary policy was tightened in 1988? Maybe that might have prevented
the inflation rate from rising too quickly to the 3 percent level at the end
of 1989. The BOJ was most likely behind the curve. However, it probably
would not have had a measurable impact on the bubble process of stock
prices and land prices. Even if the interest rate had been hiked earlier, it is
unlikely that the expected return of purchasing an asset would not have
been affected very much when the asset is in a bubble process.5
Those who emphasize the damage of burst bubble in the 1990s may argue that the mistake of monetary policy in the second half of the 1980s was
that it allowed the bubble to get bigger and bigger. There is no clear-cut answer to the question of how monetary policy should respond to asset-price
inflation with a stable CPI inflation rate, as will be seen in the general discussion in section 4.2.3.
However, the dilemma of the monetary policy at the time was that CPI
inflation was indicating low inflation, mainly due to a sharp yen appreciation, from 260 yen/dollar in February 1985 to 150 yen/dollar in the summer
of 1986, and to 120 yen/dollar in December 1987. When the CPI inflation
rate is about 0.5 percent while the stock and land prices are increasing at
30 percent annually, what should monetary policy do? The low inflation
rate, which is below typical inflation targets of around 2 percent, might
suggest there is room for monetary easing, while stopping the asset-price
inflation requires tighter monetary policy. There seems to be a dilemma for
monetary policy.
There is a fundamental law in (linear model) economics that there should
5. See Ito and Iwaisako (1996) for an interpretation of the Japanese bubble in the 1980s as
an application of stochastic bubbles. They differentiated the simulated effects of a temporary
change in the interest rate and the simulated permanent change in the interest rate upon asset prices. They argue that unless the low interest rate in the late 1980s had been perceived to
be permanent, the large increase in asset prices could not have been explained.
Two Decades of Japanese Monetary Policy and the Deflation Problem
137
be at least two policy instruments to pursue two policy objectives. No perfect solution for the interest rate policy can be obtained to pursue both CPI
price stability and asset-price stability.
Assessment of monetary policy in Japan in 1987 and 1986 is difficult.
Could one justify the monetary policy that lowered the discount rate to 2.5
percent in February 1987 and maintained it at 2.5 percent, then the record
low, until May 1989? One may argue that the BOJ should have applied tight
monetary policy in 1987 in order to curb asset-price inflation. But it would
have been difficult to justify the action given the low CPI inflation rate, the
slow economic recovery from the yen-appreciation recession of 1986, and
the aftermath of Black Monday in October 1987. We are not confident that
preventing asset-price inflation was an overriding priority of the central
bank in 1987. On the other hand, the trade-off had disappeared in 1988
when both CPI price forecasts and asset-price movements now indicated
that at least modest tightening would have been justifiable. The BOJ was
probably behind the curve in 1988.
4.2.2 Bubble Overkill?
In the beginning of the asset-price decline, public opinion was favorable
toward monetary and regulatory policy to stop the bubble. Housing was
considered to have become too expensive to ordinary citizens, so stopping
the housing price from skyrocketing was considered to be a good thing.
Despite the burst of the bubble, robust consumption and investment continued in 1991. The GDP growth rate remained higher than 3 percent in
1990 and 1991.
The Japanese economy slowed down considerably in 1992. Stock prices
plummeted in the summer of 1992, to the level of 15,000 in Nikkei 225 index, losing more than 60 percent of the peak value in two and one-half
years. The quarter-to-quarter GDP growth rate became negative in the
spring–summer of 1992. Lending to the real estate sector from banks
slowed down after 1991 due to regulation, but there was a loophole. Lending via nonbank financial institutions (such as leasing companies) continued and total lending to the real estate, construction, and nonbank sectors
remained high until the mid-1990s. Nonperforming loans, due to nonpayment of interest by real estate companies, became a popular topic of business discussion, but was not yet showing up in any banking statistics in the
first half of the 1990s.
The discount rate was lowered to 5.5 percent in July 1991, to 5 percent
in November 1991, and to 4.5 percent in December 1991. The decline of
the ODR continued in 1992 and 1993. A fiscal stimulus package was introduced in 1992 in response to the weakening economy. This was the beginning of a series of fiscal stimulus packages.
The economy was stagnant from 1992 to 1994, with the growth rate below 1.2 percent, three years in a row. Land prices continued to decline
138
Takatoshi Ito and Frederic S. Mishkin
steadily. The CPI inflation rate declined from just above 2 percent in the beginning of 1992 to 0 percent by mid-1995. Monetary policy was relaxed in
1992 and 1993 in response to weakening of the economy. The ODR was
lowered from 4.5 percent to 3.75 percent in April 1992, to 3.25 percent in
July 1992, to 2.5 percent in February 1993, and to 1.75 percent in September 1993. There was no change in the ODR in 1994, but it was lowered to 1
percent in April 1995, and finally to 0.5 percent in September 1995.
The question from the viewpoint of preventing deflation is whether or
not the pace of the interest rate cut from 1992 to 1995 was quick enough.
The fact that the economy continued to be stagnant and the inflation rate
dropped to 0 percent suggested that the BOJ might have underestimated
deflationary forces.
During the period from 1992 to 1995, the nonperforming loans problem
became worse and worse. Many construction and real estate companies
were virtually bankrupt, since the market value of real estate in inventory
had become much lower than their purchase values, and cash flows were
dwindling. As a result, these companies were having trouble making interest payments on their bank loans. However, the banks, fearing that losses
would become apparent and having a false belief the real estate market
would rebound soon, kept lending to these companies that could not service
their debt—a practice that became known as “ever-greening.” The balance
sheets of corporations and banks were quickly deteriorating.
Smaller financial institutions—housing loan companies, credit unions,
one regional bank—failed in 1995. The banking problem was worsening,
but no serious policy was introduced to address the problem. Since the seriousness was hidden behind murky accounting rules and a lenient bank
supervisor (the Ministry of Finance), the public was not informed of the
magnitude of the problem or a coming crisis. Since the public and politicians were not alarmed, there was little sympathy toward any suggestions
for fiscal injections to recapitalize the banks.
Many economists called for introducing prompt corrective action for
weak institutions and fiscal injection, if necessary, for either closing institutions or rehabilitating them. But fiscal injection was politically difficult.
Instead, in 1995, the Ministry of Finance on the one hand guaranteed all
deposits, suspending the deposit insurance ceiling, and on the other hand
declared that no major bank would fail.
In spite of a weak economy, the exchange rate was appreciating from
1993 to 1995. The exchange rate appreciated from 100 yen/dollar to 80 yen/
dollar in the spring of 1995, with no apparent macrofundamental reasons for
such a sudden move. The exchange rate appreciation dampened an expectation of early recovery and contributed to disinflation and then deflation.
The economy started to grow in the second half of 1995, and the year
1996 turned out to be a good one, with the growth rate exceeding 3 percent.
The yen depreciated to a level above 110 yen/dollar, providing additional
support for a recovery. A fragile economic recovery of 1996 accelerated in
Two Decades of Japanese Monetary Policy and the Deflation Problem
139
the first quarter of 1997, as the preannounced consumption tax rate increase of April 1997 induced consumers to accelerate big-ticket consumption. In April 1997, the consumption tax rate was raised from 3 percent to
5 percent, and the temporary special income tax cut was allowed to expire,
both as planned. The growth rate significantly slowed down in the second
half of 1997. This was the result of the Asian currency crisis, and the banking crisis of the Japanese economy in November. The economy continued
to deteriorate in 1998: the year 1998 recorded negative growth for the first
time since 1976.
From 1997 to 1998, Japanese financial markets suffered from a severe
crisis, as banks were losing capital due to high ratios of nonperforming
loans and falling asset prices. Three large banks—Hokkaido Takushoku,
Long-term Credit, and Nippon Credit—failed, and other banks were also
suffering from declining capital. Banks were curtailing lending and a severe
credit crunch was observed. The resulting negative effects on aggregate demand then pushed the economy into deflation.
The government finally decided to inject capital into the banks. The first
capital injection in March 1998 turned out to be insufficient but the second
capital injection of March 1999 finally calmed the market. Ito and Harada
(2000) showed that the Japan premium—a risk premium demanded by
western banks upon Japanese banks for interbank lending/borrowing—
disappeared after March 1999.
4.2.3 Asset Prices and Monetary Policy
In retrospect of 1985–2003, there are several questions on what the BOJ
should or could have done. The first question is whether or not the BOJ
should have prevented the bubble. If all the trouble of the 1990s originates
from the bubble, stronger actions should have been taken against the assetprice increases. This question relates to a new debate over the objective of
central banks.6 Some researchers, more than others, think that asset prices
should be considered as a part of price stability that is the sole objective of
many independent central banks. Cecchetti, et al. (2000) argued strongly
to put asset prices as direct measure of the goal of monetary policy.
Bernanke and Gertler (1999) examined monetary policy in the presence
of asset-price bubbles, with application to Japan. They built a model with
an exogenous asset-price bubble, applied alternative monetary-policy
rules, and then estimated reaction functions for the Federal Reserve (FED)
and BOJ. They applied the Clarida, Gali, and Gertler (1998) model to estimate reaction functions for the FED and the BOJ. The model assumes rational expectation for estimating expected inflation rate that is used to calculate the inflation rate gap. Their results indicate that the Japanese policy
was too tight from 1985 to 1988 and too lax from 1988 to 1990, fueling a
6. A few conference volumes dedicated to this question have been published, see for example, Hunter, Kaufman, and Pomerleano (2003); and Richards and Robinson (2003).
140
Takatoshi Ito and Frederic S. Mishkin
stock bubble, and too tight, again, from 1992 until at least 1996. They argue that even without explicitly targeting the asset prices, the BOJ should
have tightened from 1998 to 1990, probably ending the bubble much earlier.
Okina and Shiratsuka (2002) criticized Bernanke and Gertler (1999) on
the grounds that Bernanke and Gertler used a forward-looking inflation
rate as expected inflation, but the inflation rate they used was not adjusted
for consumption tax rate changes. Okina and Shiratsuka argued that the
rapid increase of interest rate derived from the policy rule of Bernanke and
Gertler mainly resulted from the introduction of the consumption tax in
April 1989. The paper by Okina, Shirakawa, and Shiratsuka (2001) contains a good review of why the bubble happened, how the BOJ reacted, and
what could have been done, from the angle of the central bank. In section
IV (Did the BOJ’s Monetary Policy Create the Bubble?) the authors take
the view that the BOJ lowered the interest rate from 1986 to 1987 to support the “policy coordination” framework, and to prevent the appreciation
of the yen. The paper then reviews the policy in 1988 and 1989.
There are many criticisms of the view that the central bank should pay
special attention to asset prices beyond their effects on CPI. See, for example, Mishkin (2001) and Mishkin and White (2003). Ito (2003) emphasizes the role of bank supervision, rather than monetary policy, for preventing a bubble or managing a burst bubble.
The difficulty in using monetary policy (raising and lowering of the interest rate) alone to prevent a bubble can be summarized as follows. First,
the central bank often would not know whether asset prices are rising due
to fundamentals or due to a bubble. Second, when the bubble is in force, it
would take a very high interest rate to pop the bubble, and that would
throw real variables into volatile fluctuations. Those skeptics emphasize
the importance of supervision policy rather than monetary policy to maintain financial stability.
Given that a bubble is created, the effects from the bursting of the bubble
could be moderated by monetary policy. The question is whether the BOJ
was behind the curve from 1992 to 1995. The BOJ may have been too slow
to ease, possibly for fear of rekindling a bubble. Similarly, the bank may
have waited too long to adopt the ZIRP, possibly because it was an unprecedented move. Would policy have been better if the bank adopted the
ZIRP earlier than February 1999?
4.3 New Bank of Japan
4.3.1 Monetary Policy of the Hayami Regime, 1998–2003
When the newly independent Bank of Japan started in April 1998, hopes
were high in that the BOJ would improve its performance and return to
what had been viewed as successful monetary policy in the preceding two
Two Decades of Japanese Monetary Policy and the Deflation Problem
141
decades. However, after five years under the Hayami regime, the BOJ had
lost credibility and suffered a serious confidence problem. What happened? The short answers to these questions are two-fold. First, the policy
board members, led by Governor Hayami, misjudged the economic conditions, maybe because they were too eager to go back to the “normal” situation where the interest rate is positive. The interest rate hike in August
2000 was a clear mistake of this kind. Second, the governor and fellow
board members took independence literally and refused to cooperate with
the government when the economic conditions called for such cooperation.
Since independence and early establishment of credibility were considered
so important, policy actions became conservative, timid, and tentative.
Cargill, Hutchison, and Ito (2000, 173) called this the “independence trap.”
Even when policy was finally directed toward quantitative easing in March
2001, this policy was not explained adequately, especially because the BOJ
had claimed that it was likely to be ineffective. Therefore, the general public
viewed the BOJ as adopting a policy that the bank did not believe in. That
was hardly a good way of communicating with the market.
The old Bank of Japan, under the 1942 Law, was supposed to pursue
monetary policy in order to maximize economic potential (not price stability), and the governor could be replaced by the minister of finance, if the
governor did not follow the government’s instructions.7 A lack of independence is often cited as a cause for an unusually high inflation rate,
about 30 percent, in 1973–74, in the wake of the first oil crisis. After the inflation of 1973–74, the BOJ had conducted prudent monetary policy,
achieving a gradual decline in the inflation rate. Cargill, Hutchison, and Ito
(1997; chap. 8) have praised the conduct of the BOJ, achieving a de facto
independence based on reputation. Japan was known to have been an
“outlier” in the relationship between the legal independence index and the
historical inflation rate.
The new law, the Bank of Japan Law of 1998, guaranteed the independence of the BOJ in its policy making and board member appointments.8
The law became effective on 1 April 1998. At around the same time, Mr.
Hayami was appointed as governor, and Mr. Yamaguchi and Mr. Fujiwara
two deputy governors. Two policy board members were carried over from
the old law regime, but four new members were appointed in April 1998 to
7. The 1942 Law specified that the BOJ conducts its operation “in order that the general
economic activities of the nation might adequately be enhanced” (Article 1). The objective of
the BOJ was “for achievement of national aims” (Article 2). These wordings should be understood in the context of the war when the bill was passed. See Cargill, Hutchinson, and Ito
(2000; chap. 4) for detailed comparison of the old and new Bank of Japan Laws.
8. The 1998 Law specifies two pillars, “the pursuit of price stability, contributing to the
sound development of the national economy (Article 2),” and “maintenance of an orderly financial system (Article 1).” The absence of mentioning full employment, economic growth,
or exchange rate objectives suggests that price stability is the primary objective. Financialsystem stability is a shared responsibility with government.
142
Takatoshi Ito and Frederic S. Mishkin
replace the old members and vacancy. Mr. Hayami, age seventy-two at the
time of new governor appointment, left the BOJ after serving for thirtyfour years on the international side of the bank, in 1981 (seventeen years
earlier) to go to a general trading company, Nissho-Iwai. After serving as
president and chairman of Nissho-Iwai, he had retired from the company
for several years, until he returned to the BOJ as governor. Deputy Governor Yamaguchi had climbed up the ladder in the BOJ with a reputation for
his knowledge about the core business of central banking. Deputy Governor Fujiwara was a former journalist. Governor Hayami was brought back
to the top position, partly because he was considered to be incorruptible in
the wake of a scandal at the Bank of Japan.9
The Japanese economy in the spring of 1998 was in the process of falling
into a serious recession and financial instability. In November 1997, financial instability became prominent: one large bank and one small bank, a
large securities firm, and a medium-size securities firm all failed, and credit
lines among the Japanese financial institutions, and between western financial institutions and Japanese financial institutions became severely
limited. The Asian financial crisis was spreading from Thailand to Indonesia, to Korea, and to the region in general. Demand was falling and it
was clear that the economy was heading into a recession.10 The overnight
call rate, the market rate corresponding to the Federal Funds rate in the
United States, at the time was about 0.4–0.5 percent. This stance was maintained until 9 September 1998, when the target of the call rate was reduced
to 0.25 percent.11
Another major step was taken on 12 February 1999. The board decided
to lower the overnight call rate as low as possible, with an immediate action
9. Many bank officials were implicated for inappropriate behavior of dining and golfing
with private-sector people. The scandal hit the media particularly hard in the first three
months of 1997. High salaries, high severance pay, and large company housing also became a
target of criticism. One bank official was arrested for taking bribes in return for leaking information to a securities firm. Governor Matsushita and Deputy Governor Fukui (who returned as governor five years later) resigned to take responsibility in March 1998, days before
the new BOJ law took effect. The official who took bribes was dismissed from the Bank on
April 3, 1998.
10. In the spring of 1998, it was announced that the economy had just experienced the two
consecutive quarters of negative growth rates: –0.7 percent in 1997:IV and –0.3 percent in
1998:I. The currently available new SNA93 (System of National Accounts, following a United
Nations recommendation of 1993) available at [http://www.esri.cao.go.jp/jp/sna/qe034-2/
gdemenuja.html] does not show this: 0.7 percent in 1997:IV and –1.0 percent in 1998:I. The
difference is due to the differences in the base year, the estimation methods, and the seasonal
adjustment method. The point is that the BOJ and the government should have had a more
negative assessment of the economy at the time of Spring 1998.
11. “The Policy Board determined to further ease the stance of money market operations
for the inter-meeting period ahead as follows: The Bank of Japan will encourage the uncollateralized overnight call rate to move on average around 0.25 percent” (Bank of Japan, Announcement of Decisions, September 9, 1998).
Two Decades of Japanese Monetary Policy and the Deflation Problem
143
to lower it to 0.15 percent.12 This is the beginning of the so-called zero interest rate policy (ZIRP). It was clear that the economy was in a very weak
state. At the time, the GDP growth rate was thought to have shrunk for five
consecutive quarters since 1997:IV.13
By the spring of 1999, the decline in economic activity became clearer—
the instability of the Japanese financial system became acute as the Longterm Credit Bank teetered on bankruptcy; bills to strengthen the financial
system were debated in the Diet; and the international financial system was
shaken by the de facto default of the Russian debts in August.14
After ZIRP was adopted, the board members were divided into three
groups, according to the disclosed minutes. Shinotsuka, who opposed
adopting ZIRP, thought that the interest rate should be raised, partly to
help pensioners. Nakahara, who had proposed lowering the interest rate
more aggressively than other members before February, frequently put forward a motion to adopt quantitative easing and inflation targeting, as actions beyond ZIRP. Both proposals were voted down with only one vote in
favor. The majority did not recognize the need to adopt any further actions
between February and September.
Since the economy was not responding to the low interest rate, the government and business sectors began to press the BOJ to adopt more aggressive quantitative easing. Just before the September 21, 1999 meeting of
the policy board, speculations were abundant in press predicting that the
policy board would adopt some sort of quantitative easing, possibly nonsterilized intervention in the foreign exchange market in cooperation with
the Ministry of Finance. The market regarded that nonsterilized intervention to be a signal that the BOJ would fight deflation with unconventional
measures. The markets also focused on whether the BOJ would increase
the amount of money market liquidity on the settlement day that was two
days after the intervention.
The policy board reacted strongly to this speculation in the press. The
board issued the statement, in addition to a brief announcement of the
monetary-policy decision, at the conclusion of the meeting, instead of
waiting for quick minutes to be released two days later. In the announce12. “The Bank of Japan will provide more ample funds and encourage the uncollateralized
overnight call rate to move as low as possible. To avoid excessive volatility in the short-term
financial markets, the Bank of Japan will, by paying due consideration to maintaining market function, initially aim to guide the above call rate to move around 0.15 percent, and subsequently induce further decline in view of the market developments” (Bank of Japan, Announcement of Decisions, February 12, 1999).
13. At the time of spring 1999, the growth rates of five quarters from 1997:IV through
1998:IV were estimated as negative. The current (spring of 2004) estimates for the same period are 0.7, –1.0, –1.1, 0.8, and 0.1. The reasons for the difference are explained in note ten.
14. Some speculate that there was also implicit political pressure from the meeting between
the Finance Minister of Japan and the U.S. Treasury secretary on September 4.
144
Takatoshi Ito and Frederic S. Mishkin
ment, the board emphasized that monetary policy would not respond to
exchange rate movements, that nonsterilized intervention was not a useful
policy, and that the press was greatly mistaken in its reports on what would
happen at the upcoming meeting. The board indicated that it had done
enough in easing monetary conditions, and it barely concealed the desire
to go back to the positive interest rate by emphasizing the side effects of
ZIRP.
The board challenged the market expectation that nonsterilized intervention was to be pursued. It took a position that the exchange rate was
one of the variables to be monitored, but monetary policy should not particularly respond to the exchange rate movement, per se.15 The board then
explained that nonsterilized intervention was not a useful concept for the
central bank that watches total funds in the market, whatever various
sources it came from.16 In addition, the board statement contained cautionary comments on the side effects of ZIRP, a forerunner to ending the
ZIRP eleven months later.17
The board expressed displeasure on press reports and market reaction in
strong words: “In the past few days, the market has substantially fluctuated
by speculations on monetary policy. What should be clear is that the conduct of monetary policy is exclusively decided by majority vote at the
Monetary Policy Meeting, a regular meeting of the Policy Board. It is never
the case that our policy is determined in advance or in consultation with
outside bodies. We would like to emphasize this point” (Bank of Japan, “On
the Current Monetary Policy” 21 September 1999).
The quotes from the statement vividly illustrated the position of the
board. Any reporting of the expected decision was considered to be a challenge to independence. The board successfully extinguished any expecta15. “The foreign exchange rate in itself is not a direct objective of monetary policy. One of
the precious lessons we learned from the experience of policy operations during the bubble
period is that, monetary policy operations linked with control of the foreign exchange rate
runs a risk of leading to erroneous policy decisions. Having said this, it does not mean that
monetary policy is pursued without any consideration to the development of the foreign exchange rate. The Bank considers it important to carefully monitor the development of the foreign exchange rate from the viewpoint of how it affects the economy and prices” (Bank of
Japan, “On the Current Monetary Policy” September 21, 1999).
16. “In relation to the foreign exchange rate policy, we have heard arguments in favor of
nonsterilized intervention. In the reserve market, however, there are various flows of funds
such as currency in circulation and Treasury funds other than those resulting from the intervention. The Bank conducts its daily market operations taking into account all the money
flows, in order to create ample reserves to such an extent as described above. This strong commitment of fund provision is consistent with the government’s current foreign exchange rate
policy” (Bank of Japan, “On the Current Monetary Policy” September 21, 1999).
17. “The Bank views the current state of the Japanese economy as having stopped deteriorating with some bright signs, though a clear and sustainable recovery of private demand has
yet to be seen. In pursuing the zero interest rate policy, we need to carefully examine its adverse side effects, but deem it important to support the economic recovery by continuing easy
monetary policy for the periods ahead” (Bank of Japan, “On the Current Monetary Policy”
September 21, 1999).
Two Decades of Japanese Monetary Policy and the Deflation Problem
145
tion in the market that the bank would be accommodative in response to
desires from the government or the market. Any doubt about independence was erased on 21 September 1999. However, their own strong words
might have trapped the board members: that is, they could not change their
positions in the following months.
Between the fall of 1999 and the summer of 2000, there was no additional easing, except for liquidity injections to deal with Y2K concerns.
The government wanted some sort of additional measures of monetary
easing, while the governor increasingly mentioned the possibility of lifting
ZIRP. At this point, the bank explained that the bank would continue
ZIRP “until deflationary concerns subside.” The economy started to show
some signs of recovery in the spring of 2000, ICT-related stock prices went
up and the Nikkei 225 increased by 30 percent between March 1999 and
March 2000. Corporate profits rose and corporate investment showed
signs of recovery. There was an argument that these corporate earnings
would trickle down to households to stimulate consumption sooner or
later.18 This argument was dubbed the “dam theory”: water was filling the
corporate dam and would overflow sooner or later. Governor Hayami, believing that this was communication with the market, frequently suggested
that there were bright signs in the economy and, as a consequence, there
would be a possibility of raising the interest rate. Critics thought it was premature to talk about lifting the interest rate, and any mention of it itself diminished the effect of ZIRP by limiting its effects through expectations that
easing would continue into the future.
The ZIRP was lifted in the policy board meeting of 11 August 2000.19 At
this point, the continuation of a recovery of the Japanese economy was at
best doubtful. First, the ICT bubble had ended and stock prices in the
United States and Japan were heading down, suggesting investment and
consumption would be adversely affected in the near future. Second, the
U.S. economy was beginning to show weakness, and Japanese exports to
the United States were expected to decline in the future. Third, the inflation rate was still negative, and there was no sign of an end to deflation.
Critics of the bank thought that ending ZIRP was a mistake. Indeed, the
government exercised an option, specified in the Bank of Japan Law, to put
18. “Currently, it is our judgment that Japan’s economy is at the stage where the number of
firms taking the offensive has started increasing, that is, the economy is moderately recovering parallel with structural adjustment. . . . with respect to the recovery of private demand, it
seems natural that the corporate sector, which has regained profitability as a result of restructuring, should take the lead by increasing investment followed by the household sector
as income conditions gradually improve. This is the development we are now witnessing”
(Speech given by Masaru Hayami, Governor of the BOJ, at the Japan Center for Economic Research on May 29, 2000, available at [http://www.boj.or.jp/en/press/00/ko0005b.htm#0103]).
19. Governor Hayami intended to raise the interest rate in July. However, a large department store, SOGO, failed and the economy showed some weakness. The plan of lifting the interest rate was postponed without being submitted to the meeting.
146
Takatoshi Ito and Frederic S. Mishkin
forward a motion for delaying voting of the proposal of raising the interest
rate until the next meeting. The government motion was overruled by the
board by an eight to one vote, and then the lifting of the zero interest rate
policy was decided by a seven to two decision.
Almost as soon as the interest rate was raised in August, the Japanese
economy entered into a recession. It was not known at the time, but the
official date for the peak of the business cycle turned out to be October
2000. The growth rate of 2000:III turned negative, which was offset to some
extent by a brief recovery in 2000:IV. But, as the economy turned into a recession, the criticism of the BOJ’s actions became stronger.
The economy weakened substantially toward the end of 2000. Many
urged changes in monetary policy. Some economists had recommended
the return to ZIRP, and others recommended quantitative easing and unconventional monetary policy including increasing the amount of regular
purchases of long-term government bonds, and newly purchasing listed
mutual funds of stocks, foreign bonds, and even real estate funds. These
unconventional monetary tools had been rejected by Bank of Japan economists earlier.
As 2001 started, many indicators were showing weakness and the Bank
of Japan decided to ease. The question then was whether to go back to the
ZIRP or to introduce a new framework, quantitative easing. In February,
the bank introduced the so-called Lombard lending facility as well as cutting the official discount rate from 0.5 percent to 0.35 percent. The Lombard lending facility was to lend automatically to banks with collateral at
the official discount rate, so that the interest rate would be capped at 0.35
percent. However, the market rate was at around 0.2–0.25 percent, so there
was little real impact from the introduction of the Lombard facility. Pressure to ease monetary conditions did not stop because of these measures in
February 2001.
The policy board meeting of 19 March 2001, turned out to be the beginning of quantitative easing as well as further easing in terms of the interest
rate. The target inter-bank rate was lowered immediately to 0.15 percent,
and would go down to zero, as conditions warranted. The official discount
rate was cut to 0.25 percent. However, the policy change was not announced
as just a return to ZIRP. It was billed as a change in the monetary policy instrument. The instrument was changed from the short-term interest rate to
the balance of current accounts at the BOJ. The target of the current account was set at 5 trillion yen. However, by targeting an amount beyond required reserves (about 4 trillion yen), it effectively meant that the interbank
rate (i.e., the call rate) would go to zero. This amounted to excess reserve
targeting.20 In September 2001, the official discount rate was cut to 0.1 percent, but this did not have any impact.
20. Earlier than it was adopted in March 2001, BOJ economist, Okina (1999a) reviewed the
excess reserve targeting as a possibility of next step of further monetary easing. He pointed
Two Decades of Japanese Monetary Policy and the Deflation Problem
147
The bank has also made clearer the conditions when it would lift ZIRP
in the future. When the BOJ adopted ZIRP for the first time in February
1999, the condition for lifting ZIRP was when deflationary concerns were
dispelled. When the ZIRP was effectively reintroduced in March 2001, the
condition became more concrete: excess reserve targeting, or de facto ZIRP,
would not be abandoned until the inflation rate, measured by CPI excluding fresh food, became stably above zero. The exit condition would be further clarified in October 2003, to be explained later.
From March 2001 to March 2003, quantitative easing was expanded in
several steps.
• In August 2001, another measure of quantitative easing was employed.
The amount of BOJ outright purchases of long-term government
bonds was raised from 400 billion yen per month to 600 billion yen per
month. At the same time, the current account target was raised to 6 trillion yen (or about 2 trillion yen excess reserves).
• In December 2001, the monthly purchase of long-term bonds was increased from 600 billion yen to 800 billion yen, the current account
target was raised to 10–15 trillion yen.
• In February 2002, the monthly purchase of long-term bonds was increased from 800 billion yen to 1 trillion yen.
• In October 2002, the monthly purchase of long-term bonds was raised
to 1.2 trillion yen from 800 billion yen, and the current account target
was raised to 15–20 trillion yen.
There have been mixed reviews on these steps. Although these steps expanded quantitative easing, especially in the amount of long-term bonds
from 400 billion yen per month in September 2001 to 1.2 trillion yen per
month in October 2002, deflation worsened. Some argue that this shows
that quantitative easing did not work. However, advocates of quantitative
easing would say that these actions prevented a major decline in economic
activities.
These measures are summarized in the figure 4.3. Panel A shows the expansion of purchase of long-term bonds and current account target, while
panel B shows the movements of the official discount rate and the call rate.
4.3.2 Assessment of the Hayami Regime
In the initial stage (April 1998 to March 1999) of the Hayami regime, until ZIRP was adopted, many BOJ officials expressed a negative view toward
further easing (zero interest rate and quantitative easing including base
money expansion, government bond, and equity purchases), indicating
that it was either ineffective or would have undesirable side effects, includout a few problems with this option. First, “what kind of function can be expected of excess
reserves” is not known with certainty and it was identified as a problem. Second, excess reserves is not reliable “as an indicator for monetary easing.” Third, Okina points out an operational hurdle.
148
Takatoshi Ito and Frederic S. Mishkin
A
B
Fig. 4.3
Quantitative easing and interest rate cut
ing the risk of high inflation.21 The call for easing by scholars was being rebuffed. (See Krugman 1998, Meltzer 1998, McKinnon and Ohno 1997 for
the calls for monetary easing; and Okina 1999a, b for the rebuff.) When
ZIRP was adopted in the spring of 1999, the bank maintained the view that
21. One such cautious opinion was expressed in July 1999 by Kazuo Ueda, a former University of Tokyo professor, now a newly appointed policy board member.
“The policy to increase the money supply would first create some decline in the call rate,
but automatically create further rate declines if the economy worsens and the demand for
money declines. In this sense the commitment to avoid deflationary forces is stronger with
money supply targeting. . . . To the extent that the money supply works through interest
rates, the commitment money supply targeting delivers is already contained in the current
policy stance. . . . The argument that an increase in the growth rate of the money supply increases inflationary expectations and stimulates aggregate demand by lowering real interest rates sounds attractive. It is unclear again, however, how this mechanism works when
the nominal interest rate has been already driven down to zero. . . . How about a policy of
letting the monetary base grow at 20 or 30 percent then? Inflation does not seem to be on
the horizon. One can tighten after the inflation rate reaches 1 or 2 percent. We think such
a policy would have a small chance of success for reasons already mentioned. When it does
Two Decades of Japanese Monetary Policy and the Deflation Problem
149
no further steps were needed. The bank strongly resented any pressure or
even suggestion from outside on further easing, as shown in the episode of
their complaining about the speculation of easing before the meeting in
September 1999. In the spring of 2000, Governor Hayami started to suggest ending ZIRP. Most likely, he wanted to communicate with the market
on the bank’s future intentions, in order to avoid a “surprise” reaction of
the market and resulting volatility in the money and capital market. However, this suggestion certainly diminished any beneficial effects of ZIRP because it created expectations of higher interest rates in the future. The interest rate was raised in August 2000 despite the opinions by many scholars
and the government of the need for further easing. In an international conference sponsored by the BOJ in July 2000, many scholars and foreign participants were critical of the past and current policy of the BOJ: Meltzer
(2001), Goodfriend (2001), and Svensson (2001)—note the publication
date of these papers was in 2001, but the conference took place in July
2000, one month before the ZIRP was reversed. Oda and Okina (2001)
compare various policy options of monetary easing and their associated
risks. The authors emphasized the risks more than the benefits of policy
options proposed to the BOJ by “academics.” They argued that “introduction of a temporary fixed exchange rate system and a huge increase in the
outright purchase of medium- and long-term government bonds can induce relatively large effects, although the uncertainty in the effects as well
as the accompanied costs and risks may be very large.” One of the discussants, Jack Beebe (2001), felt that the “authors’ views of policy feasibility
and risks are unduly pessimistic. . . . Thus, the risks inherent in taking further policy actions need to be balanced against the risks of not taking
them.” What is striking is that the conference at which this debate took
place occurred one month before the interest rate hike when ZIRP was exited, which we view was a clear policy mistake.
When the ZIRP returned with quantitative easing (current account balance of 5 trillion yen implying the excess reserve of 1 trillion yen) in March
2001, the bank did not explain why the change in policy would be effective,
and this was particularly important because the bank had not been positive on its effectiveness in the past. In the summer to fall of 2001, there were
calls for further easing by raising the current account target increase, increasing bond purchases, and purchasing equities and foreign bonds. Bank
economists were negative on these suggested actions, saying that it was impossible to raise the current balance target (no buyers of short-term paper
with zero interest rates), or no effect beyond stabilizing the financial syssucceed, it will probably generate a much higher rate of inflation than 1 or 2 percent. Because of lags in the effects of policy, the 20–30 percent money growth will continue to generate inflationary pressure even after the tightening starts” (Kazuo Ueda, Member of the
Policy Board of the Bank of Japan, at the Meeting on Economic and Financial Matters in
Kagoshima, on July 1, 1999, available at [http://www.boj.or.jp/en/press/99/ko9907a.htm]).
150
Takatoshi Ito and Frederic S. Mishkin
tem, and that risk of possible deterioration of balance sheets would be serious.22 The policy started to change in December 2001, when the current
account target was raised and long-bond purchases were raised in several
steps. What was branded impossible was now possible, and the concern
about the balance sheet, emphasized earlier by the bank itself, was buried
without addressing it formally.
In September 2002, the bank started to purchase equities that the commercial banks held but wanted to dispose of in light of declining stock
prices. Earlier, the bank had denied any possibility of purchasing stocks.23
The action was justified by the bank on the ground that it would reduce the
risk of commercial banks’ balance sheets, and it was made clear that it was
not intended as monetary policy, but rather as financial-market stabilization policy. (The decision was not made by the monetary policy board
meeting—equivalent of Federal Open Market Committee—but the regular board meeting.) However, it was not explained why the resulting risk to
the BOJ balance sheet due to financial-stabilization policy was not a big
concern, while it was for monetary policy.
In October 2000, the bank paper “On Price Stability” emphasized that
it would be difficult to focus on a particular price index as a guide to policy. Earlier, the bank was quite negative on the idea of inflation targeting.
However, in March 2001, the board decided to adopt the ZIRP plus quantitative easing until the CPI excluding fresh food showed a positive inflation rate “stably above zero.” This seemed to be a welcome switch from negative to a positive attitude toward selecting a price index and targeting a
22. “Three options for further monetary easing can be considered when money market interest rates are near zero. . . . Third, the BOJ can carry out unconventional operations by purchasing assets other than short-term Japanese government securities. . . . The third policy option is for a central bank to purchase non-traditional assets such as government bonds,
foreign currencies, corporate bonds, stocks, or real estate which are more imperfectly substitutable for base money than are short-term government securities. As stated above, central
bank operations that amount to the exchange of perfect substitutes produce little effect on
the economy. Such non-traditional operations are effective because they directly alter the
prices of the assets in question. Possible benefits and costs of this monetary policy option,
however, are extremely uncertain” (Kazuo Ueda, Member of the Policy Board, at the semiannual meeting of the Japan Society of Monetary Economics held at Fukushima University
in Fukushima City on September 29, 2001, available at [http://www.boj.or.jp/en/press/01/
ko0112a.htm#0301]).
23. Governor Hayami denied the possibility of purchasing stocks as early as 1998, and repeatedly opposed to this saying that it violates the law. “There is intrinsically a very strict limit
as to the extent to which a central bank can take on private sector risk. By shouldering such
risk and seeing a subsequent deterioration in our assets, we might lose the confidence placed
in us to fulfill our fundamental mission. Hence, the new Bank of Japan Law (effective April
1998) prohibits the Bank from purchasing equities bearing large credit and price risks. We
thus do not think it appropriate to purchase corporate debt and equity” (A summary of the
speech given by Masaru Hayami Governor, the Bank of Japan to the Kisaragi-kai meeting in
Tokyo on December 22, 1998) [http://www.boj.or.jp/en/press/98/ko9812a.htm]). The switch in
fall of 2002, why the governor changed the opinion and the purchase became possible without changing the law, was not explained.
Two Decades of Japanese Monetary Policy and the Deflation Problem
151
numerical number, but the switch was not explained. In sum, the bank has
been changing its position and action, but the switch was not explained
well, and contributed to the decline in the credibility of the bank.
4.3.3 Fukui Regime
The new Governor Toshihiko Fukui took over the leadership of the
Bank of Japan at the maturity of the five-year term of Governor Hayami in
March 2003. Two deputy governors were also replaced. One of the two new
deputy governors is Mr. Toshiro Muto who was earlier vice minister of finance; and the other Dr. Kazumasa Iwata, a former professor of economics. Iwata has been known to favor inflation targeting.
The new team moved quickly to increase the current account balance at
the BOJ. The target amount was raised from 15–20 trillion yen, at the time
of March 2003 to 30–35 trillion yen as of January 2004. The amount of
long-term bond purchases was not changed.
The biggest change has been the rhetoric. Governor Fukui has made it
explicit that the bank should maintain ZIRP until the inflation rate was
clearly above zero. He seems to indicate commitment of ZIRP into the future, a sort of commitment recommended by inflation target advocates, or
even better.24
Although the new policy is a big improvement over the last regime, there
was some room for improvement. The tolerance of inflation was not indicated with precise numbers. Therefore, it was less credible than otherwise.
One answer to such a criticism is the policy announcement of October
2003. It laid out the conditions for raising the interest rate:
First, it requires not only that the most recently published core CPI
should register a zero percent or above, but also that such tendency
should be confirmed over a few months.
Second, the bank needs to be convinced that the prospective core CPI
will not be expected to register below a zero percent. This point will be
described in such materials as the analysis and the forecasts of policy
board members in the Outlook Report. To be more specific, many policy board members need to make the forecasts that the core CPI will register above a zero percent during the forecasting period.
The above conditions are the necessary condition. There may be
cases, however, that the bank will judge it appropriate to continue with
quantitative easing even if these two conditions are fulfilled. (Bank of
Japan, monetary policy committee announcement, 13 October 2003)
24. In his speech to economists, Fukui (2003) tried to put a spin that the exit-from-ZIRP
condition that the Bank had adopted was more tolerant to inflation than a usual inflation target: “Assuming that a target has been established (for example, at 2 percent), if the expected
inflation rate rises above the target and the Bank does not start tightening at that early stage,
the actual inflation rate is likely to go beyond the target. Since the Bank’s current policy commitments do not assume such a tightening at an early stage, they actually run a risk in the direction of greater inflation than in the case of standard inflation targeting.”
152
Takatoshi Ito and Frederic S. Mishkin
Despite the good performance in the GDP growth rate in 2003:IV, the financial and capital market participants expect that ZIRP will continue for
a long time. This is a big change from the Hayami regime. So far, credibility of the BOJ to maintain ZIRP seems to be on the rise.
The recent history of Japanese monetary policy has created two basic
problems for the Japanese monetary authorities today. First, the BOJ’s
policies have left Japan in a prolonged deflationary environment in which
conventional monetary policy through lowering the short-term interest rate
is no longer effective because the policy rate has hit a floor of zero. Second,
past Japanese monetary policy, particularly under the Hayami regime, has
left the bank of Japan with a severe credibility problem in which the markets and the public are unconvinced that Japanese monetary policy can be
committed to future expansion that would return the economy to health.
Both of these problems present the bank with particular challenges in getting the economy out of deflation quickly, a subject we will return to later
in the chapter.
4.4 Why Taylor Rules Are Not the Way
to Assess Japanese Monetary Policy
In assessing the conduct of Japanese monetary policy over the last
twenty years, the following questions arise:
1. Whether the BOJ should/could have taken stronger actions against
the genesis of the bubble?
2. Whether the BOJ should/could have eased earlier in the 1992–1998
period in the aftermath of the burst bubble, in order to prevent the economy falling into deflation?
3. Whether the BOJ should/could have adopted the zero interest rate
policy earlier than February/March 1999 in its fight against deflation (as a
prevention or as a cure)?
4. Whether it was a mistake to raise the interest rate amid deflation in
August 2000?
5. Whether the bank could have pursued nonconventional policy, beyond what was actually implemented in 2001–2003 to prevent a deflationary spiral (self-fulfilling expectation) from settling in?
The Taylor rule has been a popular tool to assess the monetary-policy
stance. The nominal interest rate is regressed on the GDP gap and the deviation of the inflation rate from the target inflation rate, along with the
constant term that represents the long-run equilibrium real interest rate
and the target inflation rate. The Taylor rule might make it possible to judge
whether the monetary policy was too tight or too loose for a particular period assuming that the estimated coefficients—either for an entire sample
period or a part of it or, in some cases, from other countries—represent a
Two Decades of Japanese Monetary Policy and the Deflation Problem
153
“normal” response of the central bank. The deviation from the estimated
(fitted value) target interest rate is interpreted for too tight or too loose
monetary policy. The simplest version of the Taylor rule regression is as follows:
(1)
it r f ∗ y yt (t ∗),
where it is the policy nominal interest rate; r f is long-term equilibrium interest rate; ∗ is the target inflation rate; t is the inflation rate; and yt is the
GDP gap (the log difference of the GDP and the potential GDP). In short,
the nominal interest rate is regressed on the constant term, the GDP gap,
and the inflation deviation, and in some cases the asset prices or the exchange rate. When equation (1) is regressed for a period in which the central bank has been operated under a stable regime, then the equation with
estimated coefficients is viewed as a “reaction function” of a central bank.
Then, the fitted value of the left-hand side variable is considered to be
target interest rate (normal interest rate) that the central bank on average
would have pursued if the reaction function was followed without a deviation. If the average is interpreted as optimal, then the Taylor equation gives
the normative content. When the actual interest rate is below the target
policy rate (the fitted value) at period t, then the monetary policy is judged
to be more relaxed, compared to other periods. Then monetary policy that
is more relaxed is judged as too lax. Similarly, when the actual interest rate
is above the target rate at period t, then monetary policy at t is judged to be
too tight.
Bernanke and Gertler (1999) showed that monetary policy was too lax
in 1989–1990, and too tight from 1992 to 1996. Okina and Shiratsuka
(2002) criticized Bernanke and Gertler (1999) that their recommendations
of early tightening in the mid 1980s to prevent asset inflation were impractical. Okina and Shiratsuka think that the forward-looking inflation rate
(with rational-expectation assumption) is a source of problem.
Okina and Shiratsuka (2002, 2004) and Okina, Shirakawa, and Shiratsuka (2001) have examined monetary policy from the mid-1980s to 2002
and explored several policy options. They tend to show that monetary policy in the mid-1980s was a mistake in the sense the bubble was formed, but
monetary policy in the mid- to late-1990s was basically right, and monetary
policy after ZIRP does not have policy options.
Reifschneider and Williams (2000) quantified the effects of the zero
bound on macroeconomic stabilization capability. They argue that under a
severe contraction, open market operations alone may be insufficient to restore equilibrium. The Taylor rule should be modified to take into account
the zero bound.
Harrigan and Kuttner (2004) applied the coefficients from the United
States, simulated the path of the interest rate, and came to a conclusion:
Had the overnight rate been set according to the Fed’s policy rule, it would
154
Takatoshi Ito and Frederic S. Mishkin
have been reduced to zero by mid-1993, and remained there at least
through 1995.
Indeed, learning lessons from the Japanese situations was a popular exercise in the United States with an intention to avoid deflation. Clouse, et
al. (2000) went through a menu of options that the central bank can think
of adopting at the zero interest rate, and Ahearne et al. (2002) critically
evaluated the BOJ policy. The latter came down to a conclusion that the
Japanese monetary policy was too tight from 1992 to 1995. Bernanke and
Gertler (1999); Jinushi, Kuroki, and Miyao (2000); McCallum (2003); and
Taylor (2001) all obtained a similar conclusion that monetary loosening after 1992 was too slow (with varying changes of degree and period).
Kamada (2004) shows various estimates depending on various assumptions on output gap and data availability for decision making. Most of the
simulated results show that the target rate in 2000 remained negative, suggesting that lifting ZIRP in August 2000 was a mistake, although he refrains from such an interpretation.
Clearly, researchers have come to quite different conclusions using estimates of Taylor rules. Can this evidence be considered to be reliable? We
have our doubts. To illustrate this we estimate a regular Taylor equation to
examine the crucial assumptions and consequences. The following is the
basic data definitions: the interest rate is the call rate (collateralized call
rate until June 1985, and uncollateralized call rate after that month); the
price index is either the CPI excluding fresh food or the GDP deflator, measured as the change over the same month/quarter of previous year.
First, the GDP gap is estimated with an assumption that the potential
GDP grows with a growth rate of moving average of the past growth rates
in the sample. The potential output is further adjusted partly with the actual output with weight of 0.9 to long-run potential output and 0.1 to the
output level of t – 1:
(2)
Y ∗t (1 gt1) exp[ ln Y ∗t1 (1 ) ln Yt1 ],
where Yt–1 is the real GDP of t –1, is a set parameter of partial adjustment
and here set to be 0.9, and gt–1 is defined as
(3)
t1
1
gt1 ∑ gj .
t1 j0
Although this is an ad hoc way to define potential output, it does capture
a gradual decline in potential output in the 1990s without imposing perfect
foresight or perfect hindsight, and allowing for the possibility that the
1990s were always below potential (lost decade) rather than imposing a restriction that some years have to be above potential. As McCallum (2003)
pointed out, using a Hodrick-Prescott (HP) filter or a curve fitting method
implies that some years have to be above potential and not appropriate in
the situation that the last set of observations are suspected as being below
potential significantly. Figure 4.4 shows our estimate of GDP gap.
Two Decades of Japanese Monetary Policy and the Deflation Problem
Fig. 4.4
155
GDP gap
We will use an inflation gap that is either backward- or forward-looking.
The target inflation rate is assumed to be 2 percent. The target long-term
real interest rate is implied from the estimated constant term, namely, the
estimated constant term less 2 percent. In order to examine what the target
rate should have been for a Taylor rule ignoring the zero bound, we estimate the equations using the data from 1982:I to 1994:IV.
Table 4.1 is the result of this estimation. The upper panel is the set of estimations with backward-looking models, while the lower panel is the set
of estimations with forward-looking models. For the price index, the GDP
deflator, CPI, and CPI excluding fresh food, are used. The GDP gap is not
significant in the forward-looking regressions.
Figure 4.5 shows the target rate (depending on the forward and backward inflation rate) compared to the actual rate. The graph shows the following property. According to the graph, with an interpretation of the target interest rate (with backward inflation rate) to be a desirable rate, it can
be said that the monetary policy was too loose from 1988–89; just about
right from 1992–1995; and too loose (!) in 1996–97. However, after 1999,
that target interest rate is negative, suggesting that the zero interest rate
policy should be maintained. The Taylor rule estimates suggest lifting of
the zero interest rate toward the end of 2000, although very briefly and very
slightly above zero.25 For the reasons we outlined in the previous section,
we are doubtful about this conclusion.
Figure 4.5A, the forward-looking model, shows that the “target rate”
has been consistently above the actual rate since 1995, either with the GDP
definition or with the CPI definition. We again are suspicious of this result.
25. Ueda (2000) cited an internal study of the Taylor rule in his argument for arguing
against lifting the zero interest rate policy. See the next footnote for detailed quotes.
156
Takatoshi Ito and Frederic S. Mishkin
Table 4.1
Taylor Rule model in Japan (1982:1–1994:4)
Inflation rate gap
GDP deflator
(1)
A. Backward-Looking model
5.799***
(0.183)
GDP deflator or CPI ex fresh food
1.468***
(0.090)
GDP gap
0.214***
(0.053)
Constant
R-squared
DW
0.799
0.662
B. Forward-Looking model
5.629***
(0.323)
GDP deflator or CPI ex fresh food
1.233***
(0.110)
GDP gap
0.114
(0.081)
Constant
R-squared
DW
0.619
0.253
CPI ex fresh food
(2)
6.201***
(0.134)
1.752***
(0.133)
0.307***
(0.040)
0.791
0.323
6.067***
(0.266)
1.706***
(0.276)
0.151
(0.097)
0.594
0.162
Notes: Standard errors in parentheses are heteroskadastic-consistent.
***Significant at the 1 percent level.
The forward-looking model did not have a significant estimate of the
GDP gap.
Figure 4.5B, the backward-looking model, shows that the GDP deflator
model has had a negative target rate since 1998. However, it shows a positive target rate during the 1997–1998 period. The CPI model shows that the
target rate has become positive since 2002. Both of these results are counterintuitive.
The above results suggest that we should be quite skeptical of estimated
Taylor rules as a measure of optimal monetary-policy stance. This does not
surprise us because there are theoretical reasons for doubting the usefulness of Taylor rules to assess monetary policy, many of which have been
outlined by Kuttner and Posen (2004). First, the Taylor rule is essentially a
reaction function, and not an optimality condition. Unless the average
monetary reaction for the period of estimation is a priori known as the best
practice, one cannot interpret it as the optimum, and any deviation cannot
be evidence of too tight or too loose. Second, estimates are often quite
sensitive to the estimation period, and that is not reassuring for us to use
any particular regression results confidently without checking robustness.
Third, the output gap, an important component of the Taylor rule equa-
Two Decades of Japanese Monetary Policy and the Deflation Problem
157
A
B
Fig. 4.5 Actual versus target rate: A, backward-looking model; B, forwardlooking model
tion, is difficult to estimate. Fitting a linear trend or more sophisticated
curve (e.g., HP filter) is unlikely to give us a correct output gap if years near
the end of the sample are unusual (either in the upward or downward direction). For example, during the late 1990s, any conventional measure of
GDP gap (or not-accelerating-inflation-rate unemployment [NAIRU]) in
the United States was indicating an overheating that would require monetary tightening. However, in view of strong productivity increase, later
known as a new economy, the Federal Reserve did not tighten monetary
policy, and strong economic growth was extended without inflation until
2000. This episode shows the difficulty in estimating mechanically the output gap. Fourth, there are some deep conceptual problems in even deciding
what an appropriate measure of the output gap is. Fifth, the regular Taylorrule estimation does not assume that the nominal interest rate is bounded
at zero. Therefore, in the case of Japan, the target rate estimated from the
Taylor rule, without imposing the zero bound, often shows that the rate
158
Takatoshi Ito and Frederic S. Mishkin
should be negative—this is regarded as evidence for keeping the rate at
zero, as illustrated above.26 However, this is not rigorous. If the Taylor rule
is to be interpreted as an optimal monetary policy, the zero constraint of the
nominal interest rate should be included in the estimation itself. Policy advice that the interest rate should be lowered more quickly than when the inflation rate is very low cannot be obtained from a regular Taylor equation
that does not impose the zero bound condition.
Our examination of the Taylor-rule exercise above leaves us with a skeptical view of the use of this model to assess Japanese monetary policy. An
alternative approach that we find very attractive is that in Kuttner and
Posen (2004), which looks for deflation scares—sharp declines in longterm bond prices when there was no increase in the short-term policy rates.
Indeed, Kuttner and Posen (2004) come to a similar view of the Hayami
regime. Monetary policy during that period weakened the credibility of the
BOJ to overcome the deflationary environment, and the abandonment of
ZIRP in August 2000 was a clear policy mistake that led to entrenched expectations of continuing deflation.
4.5 How Costly is Deflation?
We have seen that Japan’s deflation has been accompanied by weakness
in the economy. However, does this mean that deflation has been harmful in
Japan? Furthermore, even if deflation has had serious negative consequences for Japan, does that mean that deflation is always costly? When is a
deflation likely to be harmful and thus to be avoided, and when not? There
are several potential costs to deflation and we look at each of these in turn.
4.5.1 Deflation and the Labor Market
One argument for a high cost to deflation is found in the work of Akerlof,
Dickens, and Perry (1996). Inflation that is at too low a level (which for
them is below 2 percent) produces inefficiency and will result in an increase
in the natural rate of unemployment. They argue that downward rigidity of
nominal wages, which they argue is consistent with the evidence, indicates
that reductions of real wages can occur only through inflation. The implication is that a very low rate of inflation might prevent real wages from ad26. An interpretation of negative target rate under the Taylor rule as a suggestion for keeping the zero interest rate policy in Japan has been mentioned in Ueda (September 22, 2000)
who had voted against lifting ZIRP a month earlier: “the [output] gap is larger, on the deflationary side, by about 4 percent than the neutral level. With a coefficient of 50 percent on the
gap in the Taylor rule formula, the gap term already contributes –2 percent to the interest rate.
The inflation term also contributes negatively . . . Thus, there is no chance for the Taylor rule
rate to become positive under such assumptions.” A similar view was expressed by “one member” in the Monetary Policy Board meeting that reinstated ZIRP on March 19, 2001: “This
member said that, according to this member’s simulation applying the Taylor rule, when the
economy recovered in the future, termination of the policy when the inflation rate was slightly
above zero percent would not be premature.”
Two Decades of Japanese Monetary Policy and the Deflation Problem
159
justing downward in response to declining labor demand in certain industries or regions, thereby leading to increased unemployment and hindering
the reallocation of labor from declining sectors to expanding sectors.
The evidence for the Akerlof-Dickens-Perry mechanism through which
low inflation raises the natural rate of unemployment is not at all clear-cut
(e.g., Lebow, Stockton, and Wascher 1995; Card and Hyslop 1997; Lebow,
Saks, and Wilson 1999; Crawford and Harrison 2000; and Fares and Lemieux 2000). Also as pointed out by Groshen and Schweitzer (1996, 1999),
inflation can not only put “grease” in the labor markets and allow downward shifts in real wages in response to a decline in demand along the lines
of Akerlof, Dickens, and Perry (1996), but it can also put in “sand” by increasing the noise in relative real wages. This noise reduces the information
content of nominal wages and hence the efficiency of the process by which
workers are allocated across occupations and industries. Thus, we do have
some skepticism about the Akerlof-Dickens-Perry (1996) argument which
argues for keeping the inflation rate above 2 percent. However, their work
does suggest that deflation might be costly.
In Japan, downward rigidity of annual compensation may not be large
since regular, full-time workers have several months of compensations in
bonuses. The bonuses are known to be more flexible, reflecting the performance of companies as well as individuals. Kuroda and Yamamoto (2003a,
b) argued that the impact wage rigidity has on unemployment is quite small
in Japan, at least among regular male workers. Wage rigidity was found to
be more prominent among hourly-wage, part-time female employees.
Kuroda and Yamamoto (2003c) conducted a simulation analysis to show
that the downward rigidity would raise the unemployment rate by as much
as 1.8 percentage points under the baseline parameters. The downward
wage rigidity affects the labor-market condition most for the inflation rates
between 2.4 percent and 1 percent.
Whether due to the wage rigidity or to some other reasons, the unemployment rate became as high as 5.5 percent in August 2002, compared to 4
percent in April 1998. It appears that the Phillips curve in Japan is sharply
kinked at around the zero percent, CPI inflation rate.
4.5.2 Deflation, Wealth Redistribution, and Financial Instability
Unexpected deflation has the effect of shifting resources from borrowers
to lenders when there are long-term debt contracts with fixed nominal interest rates. With a lower price level, and debt fixed in nominal terms, the
real burden of this debt necessarily increases. One might think that losses
by borrowers would be offset by gains to lenders in the macro sense, since
unexpected deflation is just a wealth transfer, or a zero-sum result. But, this
is not the case because deflation can lead to financial instability which can
impose large costs on the economy. This provides an even more compelling
reason to worry about deflation.
The transfer of resources from debtors as a result of deflation means that
160
Takatoshi Ito and Frederic S. Mishkin
they suffer a loss of net worth and a deterioration in their balance sheets.
Irving Fisher (1933) aptly named this phenomenon “debt deflation” and
saw it as a major factor promoting the economic downturn during the
Great Depression.27 With less net worth, there is a decline in the amount of
collateral a lender can grab if the borrower’s investments turn sour, and the
reduction in collateral therefore increases the consequences of adverse selection because, in the case of a default, losses from loans are likely to be
more severe. In addition, the decline in net worth increases moral hazard
incentives for borrowers to take on excessive risk because they now have
less to lose if their investments go sour. The increase in moral hazard and
adverse selection from deflation then means that the financial system markets will no longer be as capable of allocating capital to productive uses,
with the result that investment will decline and the economy will contract.
Wealth transfers are thus not neutral because they interfere with the effective functioning of the capital markets. The Great Depression is an example of when deflation had very negative consequences for the economy
(Bernanke 1983; Mishkin 1978, 1991, 1997), with a recent example being
that of Japan (Mishkin 1998).
Wealth redistribution from deflation also affects the fiscal position of the
government. One of the largest borrowers with fixed interest rate is the
Japanese government. The Japanese government has been regularly issuing long-term government bonds with fixed exchange rate. (Only in 2003,
did the Japanese government start to issue inflation-indexed bonds, but the
principal is protected from deflation.) Unexpected deflation during the
1990s meant that the Japanese government had an increased debt burden
in real terms. In addition, since tax brackets are not adjusted for inflation,
deflation meant that the government had less tax revenues (i.e., it suffered
a reverse bracket creep).
4.5.3 Deflation, the Zero Bound for Nominal Interest Rates, and
Increasing Difficulties in Conducting Monetary Policy
When the economy falls into deflation, as it has in Japan recently and as
occurred in the Great Depression in the 1930s in the United States, there is
a problem that arises from the zero bound of nominal interest rate. Lenders
will not accept a negative interest rate, since hoarding cash provides a
higher return. Thus nominal interest rates cannot go below a floor of zero
and this can throw the economy into a disequilibrium situation.
Suppose that the economy is extremely weak and the real interest rate
should be very low, possibly even zero or negative, in order to stimulate a
recovery. However, when deflation is under way so that expected deflation
27. One might think that when both general price levels (for example, CPI) and asset-price
levels decrease, the ratio of asset prices to CPI may not drop as much, and the debt deflation
may not be so acute. However, the decrease in asset prices in Japan far exceeded CPI changes,
so debt deflation was a real problem.
Two Decades of Japanese Monetary Policy and the Deflation Problem
161
is substantial, a nominal interest rate that has hit the floor at zero leaves the
real interest rate quite positive. Because of the zero bound, monetary policy is no longer able to get the real interest rate down to the equilibrium real
rate that will put the economy back on track. The economy can be described as being in a “deflation trap” in which it operates below capacity
with investment discouraged due to the deflationary environment.
Summers (1991) and a board member of the BOJ (Ueda 1999) have argued that in this situation, monetary policy becomes ineffective.28 However, we believe this argument is a fallacy for the reasons outlined in
Meltzer (1995) and in Mishkin (1996), and we discuss this further below.
Monetary policy works through many other asset prices besides those of
short-term debt securities, and so even when short-term interest rates hit
the floor of zero, monetary policy can still be effective, and indeed was so
during the Great Depression (see, Romer 1992).
Nonetheless, monetary policy becomes more difficult during deflationary episodes when interest rates hit a floor of zero because the usual guides
to the conduct of monetary policy are no longer relevant. In recent years,
much of the research on how central banks should optimally conduct monetary policy focus on so-called Taylor rules, in which the central bank sets
the short-term interest rates at a level which depends on both output and
inflation gaps. The Taylor (1999) volume is an excellent example of this
type of research. However, once the interest rate hits a floor of zero, all of
the research on optimal monetary-policy rules represented by work of the
type in the Taylor (1999) volume is no longer useful because manipulating
short-term interest rates is no longer an effective tool of monetary policy,
as explained in an earlier section. We will see below that monetary policy
can still be effective in stimulating the economy, but central bankers now
will find themselves at sea without the usual knowledge to guide them,
making it harder for them to get monetary policy exactly right.
4.5.4 Productivity-Driven Deflation
There may be one type of deflation that is not necessarily harmful to the
economy: when the deflation occurs as a result of an extremely favorable
productivity shock. In this case, the debt-deflation phenomenon may not
operate. Think of what happens to a firm which finds the prices of the
goods it produces falling because of a favorable productivity shock. It is
true that the real indebtedness of the firm in terms of the firm’s good prices
28. “Now let me briefly touch upon an academic, not a real-world, question of what a central bank can do beyond zero rates if it ever wanted to ease from that point on. . . . discussing
money supply effects on the economy other than through interest rates. I must say they are
very small once liquidity has been injected enough to maintain the zero rate. . . . I hasten to
add that, once the zero rate is reached and spreads over to most of the short-term interest
rates, attempts to expand the money supply themselves may become unsuccessful. We have
been experiencing this lately in Japan” Ueda (1999).
162
Takatoshi Ito and Frederic S. Mishkin
rises. However, the value of the real value of the assets will also rise by the
same proportion, because the firm has become more productive. In this
case, the deflation is not leading to a decline in net worth and thus does not
lead to the negative consequences we described earlier. This may explain
the results in Atkeson and Kehoe (2004), who find that deflation was only
clearly associated with economic depressions during the 1930s. Indeed, we
have seen a recent episode of deflation which does not appear to have had
negative consequences for the economy, China from 1997–2003.
Also, deflation that results from favorable supply shocks may not create
a problem for the conduct of monetary policy. Favorable supply shocks are
likely to increase the productivity of capital and thus raise the natural real
rate of interest. Thus, even with deflation, the zero lower bound for interest rates will not be binding and monetary policy can be conducted using
the conventional interest-rate tools.
Deflation (or disinflation) due to productivity increases would be accompanied by faster growth of output. This is likely to be what happened in the
United States in the second half of the 1990s with the advent of the new
economy. However, deflation driven by productivity growth does not describe the situation in Japan where stagnation has accompanied deflation.29
4.5.5 Bottom Line on the Costs of Deflation
The conclusion here is that deflation that occurs as a result of a decline
in aggregate demand is likely to be harmful, both because it interferes with
the efficient functioning of the financial markets, but also because it makes
monetary policy harder to conduct. This is exactly the situation which
Japan has been experiencing recently and which the world faced during the
Great Depression period of the 1930s. This provides an important rationale for being concerned about the possibility of deflation. However, deflation which results from favorable supply shocks may not be nearly as harmful to the economy.
4.6 Deflation Prevention
The experience in Japan as well as the analysis in the previous section
suggests that deflation can be a serious problem with high costs to the
economy, particularly when it leads to a deflation trap in which conventional monetary policy is unable to help the economy to recover. Here we
examine the question of how can monetary policy be designed to prevent
deflation and a deflation trap from occurring. We wait until the next section
to explore what can be done to get out of a deflation trap once it occurs.
Clearly, as Ahearne, et al. (2002) have pointed out, one way to prevent
29. Repeated reference to “supply-side factors” and “technological innovations” in
Hayami’s speeches in the early years was quite puzzling to say the least.
Two Decades of Japanese Monetary Policy and the Deflation Problem
163
deflation is for monetary policy to respond particularly aggressively when
negative shocks hit the economy if the economy is already in a low inflation
environment. Indeed, our discussion of the Japanese experience earlier in
the chapter has shown that the BOJ did not do this and was continually behind the curve in easing monetary policy when deflationary shocks first hit
the economy. At one point, the BOJ even raised the interest rate amid deflation. Clearly, central bankers are human and do make mistakes, but can
monetary policy be designed so that deflation and deflation traps are less
likely?
Here we will see that putting in place a strong nominal anchor through
an inflation targeting regime is an important strategy for reducing the
probability that deflation will occur. However, a key issue for such an inflation targeting regime is what is the optimal level of inflation for the target? Once we examine this issue, we go on to look at whether or not it would
be better to have the inflation targeting regime shoot for an inflation target
or a price level target.
4.6.1 Inflation Targeting
As discussed and outlined in Mishkin (1999a), an inflation targeting
regime involves five elements: (a) public announcement of medium-term
numerical targets for the price-level path or inflation;30 (b) an institutional
commitment to price stability as the primary, long-run goal of monetary
policy and to achievement of the price stability goal; (c) an information inclusive strategy, with a reduced role for intermediate targets, such as money
growth; (d) increased transparency of the monetary-policy strategy
through communication with the public and the markets about the plans
and objectives of monetary policymakers; and (e) increased accountability
of the central bank for attaining its inflation objectives.
Two features of an inflation-targeting regime can help in the prevention
of deflation. The fact that a central bank that announces an inflation target and is accountable for achieving this target means that it will be under
greater pressure to take steps to avoid a deflation as long as the inflation
target is not too low (something that we turn to shortly).
For example, consider what might have happened if the BOJ had an inflation target of 2 percent for the CPI (the median for inflation-targeting
regimes) in 1992 when the CPI inflation rate was still above 2 percent.
Would that have helped the BOJ guide its policy in prevention of deflation?
Or, suppose that the BOJ was given a 2 percent inflation target, in contrast
to the actual 0.3 percent inflation rate, as well as instrument independence
30. To date all inflation targeters have chosen to target an inflation rate rather than the price
level. However, logically an inflation targeter could just as easily choose to target the path for
the price level, which trends upward at a chosen inflation rate as it targets a particular rate of
inflation itself. The only difference is whether by-gones are allowed to be by-gones. We look
at the question of the desirability of price level versus an inflation target later.
164
Takatoshi Ito and Frederic S. Mishkin
in 1998 under the new BOJ law. Would that have made the BOJ introduce
the zero interest rate policy earlier than March 1999 and avoided lifting it
in August 2000?
The inflation numbers that came in after adoption of inflation targeting
would have indicated that the BOJ was not meeting its goals and pressure
on the BOJ to pursue more expansionary monetary policy would have
clearly increased. The likelihood that the BOJ would have lowered interest
rates more rapidly and started the zero interest rate policy (ZIRP) earlier
than 1999 would have been high. Furthermore, it is almost inconceivable
that the BOJ would have abandoned ZIRP and raised the policy interest
rate in August 2000 if an inflation-targeting regime of this type had been in
place, since it was absolutely clear at the time that deflation was ongoing
and a 2 percent inflation rate was nowhere in sight. An inflation-targeting
regime is thus likely to have prevented BOJ’s mistakes after 1998 and monetary policy would have moved in the right direction far earlier. Although
this counterfactual does not imply that deflation would have been avoided,
the simulations in Ahearne et al. (2002) suggest that easing monetary policy earlier and not tightening in 2000 would have promoted a stronger
economy and reduced the degree of deflation. If an inflation-targeting
regime and operational independence of the BOJ had been in place after
the bubble economy burst, there is even a possibility that deflation could
have been avoided altogether because the BOJ would have been under continual pressure not to get behind the curve as it did in the 1992 to 1998 period when inflation was clearly below 2 percent.
The second feature of inflation targeting is that it necessarily focuses on
the management of expectations, which is increasingly viewed as being crucial to the successful conduct of monetary policy. One consequence of the
adoption of inflation-targeting regimes is that it puts in place a strong nominal anchor that helps pin down inflation expectations (e.g., see Erceg and
Levin 2001). Modern monetary theory (see Woodford 2003) shows that a
strong nominal anchor that pins down inflation expectations has major
consequences for the path of actual inflation and makes deflation much less
likely. These theoretical results are borne out by recent experience where
we have seen major successes in the ability of monetary policy to control
inflation in many industrialized countries. We would argue that this is not
because central banks have become so much more knowledgeable about
the transmission mechanisms of monetary policy. What has changed in recent years is that central banks in industrialized countries have been able
to put much stronger nominal anchors in place. The result is greatly improved performance on both the inflation and output fronts. This of course
has been done by adoption of inflation targets, as in New Zealand, Canada,
the United Kingdom, Sweden, and Australia, and to some extent in the European Monetary Union.
However, a strong nominal anchor can be put into place without a for-
Two Decades of Japanese Monetary Policy and the Deflation Problem
165
mal inflation target through direct communication with the public about
the commitment to price stability and actions that are consistent with it.
This is the strategy pursued by the Federal Reserve, which has as strong a
nominal anchor as inflation-targeting central banks, although it is embodied in an individual, Alan Greenspan (Mishkin 2000).31 This has worked
well in the United States because Greenspan has understood and emphasized in his testimony and speeches that a central bank should be highly
concerned about preventing deflation (Greenspan 2002, 2004). However, it
can be dangerous to rely on an individual to do the right thing.
Governor Hayami clearly did not understand the dangers of deflation
and continually spoke about the dangers of inflation even when the problem
for Japan was the opposite. Furthermore, as we have seen, the BOJ’s actions
under Hayami were not oriented to preventing deflation. As a result, the
BOJ has had a credibility problem, particularly under the Hayami regime,
in which the markets and the public did not expect that the BOJ to pursue
expansionary monetary policy in the future, which would ensure that deflation would end. These mistakes in the management of expectations are a
key reason why Japan found itself in a deflation that it is finding very difficult to get out of.32 Indeed, one of the reasons that one of us has advocated
inflation targeting for the United States is that an institutional basis for the
nominal anchor is likely to remain strong regardless of who is the head of
the central bank (Mishkin 1999a; Mishkin and Posen 1997; Bernanke,
Laubach, Mishkin, and Posen 1999; Mishkin 2004a, b). In the case of
Japan, having an inflation-targeting regime would have made if far more
likely that expectations would have been managed more to prevent deflation, both through actions and words, as advocated by one of us, Ito (1999).
Earlier suggestions for inflation targeting were made to help raise inflation expectation in order to get out of the deflationary trap (see Krugman
1998). Advocates of inflation targeting also suggested that it would be an
appropriate monetary-policy framework for an independent central bank
in order to enhance accountability and transparency of its policy. Ito
(1999) further argued that inflation targeting probably enhances instrument independence. Cargill, Hutchison, and Ito (2000; chap. 5) and Ito
and Hayashi (2004; chap. 5) also review major issues in the debate on inflation targeting in Japan.
31. This does not mean that there aren’t reasons for the Federal Reserve to move to an inflation target. See Mishkin (2004).
32. BOJ officials have been quite skeptical of their ability to influence inflation expectation
of the public. “The argument that an increase in the growth rate of the money supply increases
inflationary expectations and stimulates aggregate demand by lowering real interest rates
sounds attractive. It is unclear again, however, how this mechanism works when the nominal
interest rate has been already driven down to zero” (Kazuo Ueda, “The Bank of Japan’s Forward Looking Approach”—Remarks by Kazuo Ueda, member of the policy board of the
Bank of Japan, at the Meeting on Economic and Financial Matters in Kagoshima, on July 1,
1999).
166
Takatoshi Ito and Frederic S. Mishkin
The BOJ was not warm to inflation targeting. Many policy board members as well as staff economists expressed skeptical views in recorded minutes of the policy board meetings (see Fujiki, Okina, and Shiratsuka [2001];
Okina [1999a, b]; and Ueda [2000] for a succinct view). The skeptics argued
that there was no credible tool, beyond ZIRP, to raise the inflation rate. Inflation expectations in the market would not respond to a mere announcement of the target. Therefore, committing to a target when the bank did not
have the tools to achieve it would cause the bank to lose credibility.33
At the earlier stage, that is 1999–2000, there was also an argument that
the definition of deflation was not clear: which prices should be used and
what numbers should be looked at in defining deflation.34 The BOJ was also
responding to new calls for more careful definitions of price stability. On
13 October 2000, two months after raising interest rates, the policy board
issued a report called “On Price Stability.” In the document, price stability
was defined as a state that is neither deflation nor inflation. Its apparent
tautology did not help settle the problem.
Only in March 2001 did the BOJ identify the price index relevant in policy discussions as the CPI index excluding fresh food (CPIexFood).35 The
relaxed monetary policy would continue until inflation rate measured by
the CPIexFood would become stably above zero. In October 2003, “stably”
was further defined as above zero for a few months and when there would
be no risk of falling back into deflation.
It is not immediately clear to us why the BOJ was so negative toward nonconventional monetary policy and inflation targeting under the Hayami
regime.36 One possible answer was that inflation targeting was interpreted
as a strategy to inflate away the nonperforming loans problem. Governor
Hayami repeatedly cautioned that economic boom and inflation would
make problem firms survive longer: inflation would delay structural reform.37 This smacked of the view that “cleansing” was needed, which has a
33. “[T]he BOJ argues, as is recorded in the minutes of Monetary Policy Meetings, that
‘since we cannot explicitly show the way to achieve the desired inflation rate, such action
would most likely result in the BOJ losing credibility’” (Okina 1999b, 165). Critics argued that
there are nonconventional monetary-policy measures that surely make the inflation rate go
from negative to positive, the credibility argument is based on incorrect assumptions.
34. “Price indicators such as the GDP deflator, CPI, and Wholesale Price Index (WPI) often move differently. Even when these indicators exhibit the same movement, the extent to
which the sound development of the national economy will be achieved may depend on such
factors as whether property prices are stable or rising sharply” (Okina 1999a, 164).
35. The Cabinet Office also changed the definition of deflation in the appendix to the
monthly report of March 2001 from “a state in which prices are falling while the economy is
contracting” to “a state of continuing fall of prices” Available at [http://www5.cao.go.jp/
keizai3/2001/0316getsurei/main.html]).
36. Ito (2004) examines why the BOJ did not adopt inflation targeting, based mostly on
minutes of the monetary policy meetings.
37. “When the economy recovers, as is now happening, it might well be the case that efforts
for structural reform might be neglected due to a sense of security. In addition, when the
shadow of structural reform becomes conspicuous, for example in employment, calls to re-
Two Decades of Japanese Monetary Policy and the Deflation Problem
167
strong resonance to what Federal Reserve officials said during the Great
Depression in the 1930s. This view clearly misinterpreted what inflation
targeting is about.
Second, another possibility is that the BOJ fell into the “independence
trap,” as it was called by Cargill, Hutchison, and Ito (2000). Namely, the
BOJ was afraid to take bold actions when it had just gained independence.
Before independence, a usual argument not to lower the interest rate
quickly was that once it was lowered, it would be very difficult, politically,
to raise the interest rate. Achieving independence was supposed to solve this
problem. Flexible adjustments and bold actions were supposed to have become possible. On the contrary, the BOJ became much more conservative
in the sense that it became reluctant to take actions, especially unprecedented ones, that might be judged a failure later, arguing that it would be
important to establish credibility early. If this is the case, the BOJ was given
independence precisely at the moment that it should not be given independence, because the economy called for unprecedented monetary policy.
Third, one more possible interpretation is that the bank genuinely was
worried about possible deterioration of its balance sheet. Purchasing a
large amount of long-term government bonds would put the balance sheet
at risk if they later declined in value. A question is whether stopping nonconventional monetary policy on the grounds of a concern about the balance sheet is desirable from the point of view of avoiding deflation and
maximizing potential output. The BOJ is part of the public sector, and any
losses on the bank’s balance sheet would be counterbalanced by gains on
the central government’s balance sheet. Since the BOJ should be considered as a part of the government from an accounting point of view, concern
about these losses is unwarranted, unless they created political problems
for the bank. The balance sheet of the BOJ should be guaranteed by the
government if it makes sense for the BOJ to take risk in its operations.38 In
this sense, independence came at a wrong moment in history.
4.6.2 What is the Optimal Level of Inflation?
A key issue in any inflation-targeting regime, whether it targets a path of
the price level or the inflation rate, is what is the optimal level of inflation
verse such reform and pressure for additional macroeconomic policy measures such as the expansion of aggregate demand are very likely to intensify. . . . Structural problems cannot be
solved solely by macroeconomic policy measures such as monetary and fiscal policy. Now
that financial and capital markets are highly globalized, any attempt to wipe out past problems by generating inflation will never be successful.” (Speech given by Masaru Hayami, Governor of the Bank of Japan, at the Japan Center for Economic Research, Available at [http://
www.boj.or.jp/en/press/00/ko0005b.htm] May 29, 2000)
38. Under the old BOJ law, before 1998, heavy losses on the balance sheet incurred by the
BOJ were automatically filled by the Ministry of Finance. In the new law of 1998, since policies of the BOJ were subject to direction of the Minister of Finance, the clause was eliminated
that emphasizes independence of the BOJ.
168
Takatoshi Ito and Frederic S. Mishkin
that the central bank should want the price level to grow to over the long
run? In order to decide on the appropriate long-run inflation goal, we need
to answer the deeper question of what does price stability mean? Alan
Greenspan has provided a widely cited definition of price stability as a rate
of inflation that is sufficiently low so that households and businesses do not
have to take it into account in making everyday decisions. This definition
of price stability is a reasonable one and operationally, any inflation number between 0 and 3 percent seems to meet this criterion. Some economists,
Martin Feldstein (1997) and William Poole (1999) being prominent examples, argue for a long-run inflation goal of 0 percent, which has the psychological appeal of the “magic number” of zero. Indeed one concern is
that an inflation goal greater than zero might lead to a decline in central
bank credibility and instability in inflation expectations, which could lead
to an upward creep in inflation. However, evidence in Bernanke, Laubach,
Mishkin, and Posen (1999) suggests that maintaining a target for inflation
above zero, but not too far above (less than 3 percent), for an extended period, does not lead to instability in the public’s inflation expectations or to
a decline in central bank credibility.
The BOJ (2000) attempted to define price stability in October 2000.
However, it concluded that it would not be appropriate to give a numerical
value to price stability, a surprisingly negative attitude toward commitment
to inflation targeting:
If some numerical values are adopted as the definition of price stability,
they are expected to be valid for a very long period of time. In view of the
current development of prices in Japan, it is difficult to set specific numerical values to the definition of price stability that are consistent with
the sound development of the economy. Furthermore, even if some numerical values were announced, they would not serve as a reliable guidepost in the conduct of monetary policy, and the exercise would not likely
contribute to enhancing transparency of the conduct of monetary policy. Therefore, it is not deemed appropriate to define price stability by
numerical values. (Bank of Japan, 2000, Summary, Paragraph 5 [2])
There are several reasons why the desirable target rate should be positive. First, there is an upward bias in CPI by construction.39 Second, it helps
the economy to achieve necessary relative price adjustment if some prices
and wages are sticky downward. This is a basis of the argument in Akerlof, Dickens, and Perry (1996). Third, an even more persuasive argument
against an inflation goal of zero is that it makes it more likely that the economy will experience episodes of deflation. We have argued above that de39. The Laspayres index tends to underestimate the true inflation by keeping the basket
fixed, so that the demand shift due to the relative price changes would not be reflected. The
new products would not be included. Quality improvement is often ignored.
Two Decades of Japanese Monetary Policy and the Deflation Problem
169
flation can be highly dangerous when it promotes financial instability and
in addition can make monetary-policy decisions harder if as a result shortterm interest rates hit a floor of zero. The implication is that undershooting a zero inflation target (i.e., a deflation) is potentially more costly than
overshooting a zero target by the same amount. This can be dealt with by
having a target rate with a buffer so that even some perturbation around
the target would not force the economy into deflation.
The logic of this argument suggests that setting an inflation target a little
above zero is worthwhile because it provides some insurance against
episodes of deflation. Simulation evidence in Fuhrer and Madigan (1997);
Orphanides and Weiland (1998); and Reifschneider and Williams (2000)
bear this out, finding that inflation targets near zero (below 2 percent) increase output variability. This is why one of us has argued in Mishkin
(1999a) and Bernanke, Laubach, Mishkin, and Posen (1999) for a long-run
inflation goal of 1 percent above true inflation. The Boskin commission
(Boskin, et al. 1996) estimated that the measurement bias in CPI inflation
was about 1 percent and this is why Bernanke et al. (1999) suggested a CPI
inflation goal for the United States of 2 percent.40 In the case of Japan, the
upward bias in measured CPI inflation over true inflation has been estimated to be 0.9 percent (Shiratsuka 1999), although redefinition of the
price CPI in Japan may mean that the bias is now lower. Adding this to an
inflation goal of about 1 percent true inflation, an inflation goal of near 2
percent for the CPI in Japan makes sense.
Another reason why central banks might be better off with a long-run
inflation goal above zero, is that it is crucial that they not be perceived as
being overly obsessed with controlling inflation at the expense of output
stability. If a central bank is perceived as an “inflation nutter” in Mervyn
King’s (1996) terminology, in which the central bank puts no weight on
output fluctuations in making its decisions about monetary policy, it is
likely to lose the support of the public. Too low an inflation target may signal to the public that the central bank does not care sufficiently about the
public’s concerns. It is unstable for a central bank in a democracy to have
a very different loss function than the public (Blinder 1998, and Mishkin
1999b), and pursuing too low an inflation target may weaken the support
for central bank independence.41
40. Since the Boskin commission, the Bureau of Labor Statistics (BLS) has altered its procedures to reduce the measurement bias in CPI inflation. Also the inflation bias in the prices
of consumer expenditure (PCE) deflator, which appears to be the preferred measure of inflation used by the Federal Reserve, is even lower than for the CPI. This would suggest an even
lower inflation goal for this deflator.
41. By also emphasizing that the horizon for hitting an inflation target will need to be
lengthened in order to not impose large output losses on the economy if inflation is far from
target, the central bank can also make clear that it does put a weight on output fluctuations
in making its decisions about monetary policy. See Mishkin (2004).
170
Takatoshi Ito and Frederic S. Mishkin
4.6.3 Price-Level Versus Inflation Target?
Currently, all countries who have adopted inflation targeting have chosen to target inflation rate rather than the price level. However, which of
these two targets would result in better economic performance and prevent
deflations is still an open question. Here we look at this question when
the economy is assumed to be experiencing a positive rate of inflation. The
answer to this question may be quite different when the economy is already
in a prolonged deflation and we will address this situation in the following
section.
There are two key advantages of a price-level target relative to an inflation target. The first is that a price-level target can reduce the uncertainty
about where the price level will be over long horizons. With an inflation target, misses of the inflation target are not reversed by the central bank. The
result is that inflation will be a stationary stochastic process, that is, integrated of order zero, I(0), while the price level will be nonstationary, an I(1)
process. The result is that the uncertainty of where the price level will be in
the future grows with the forecast horizon. This uncertainty can make
long-run planning difficult and may therefore lead to a decrease in economic efficiency. Although, McCallum (1999) has argued that the amount
of long-run uncertainty about the future price level that would arise from
successful adherence to an inflation target may not be all that large, it still
complicates the planning process and may lead to more mistakes in investment decisions.
The second possible advantage of a price-level target is that in models
with a high degree of forward-looking behavior on the part of economic
agents (e.g., Svensson 1999; Woodford 1999, 2003; Svensson and Woodford 2003; Clarida, Gali, and Gertler 1999; Dittmar and Gavin 2000; Vestin
2000) it produces less output variance than an inflation target. However,
empirical evidence (Fuhrer 1997) does not clearly support forward-looking
expectations formation, and models with forward-looking behavior have
counterintuitive properties that seem to be inconsistent with inflation dynamics (Estrella and Fuhrer 1998).
The traditional view, forcefully articulated by Fischer (1994), argues that
a price-level target might produce more output variability than an inflation
target because unanticipated shocks to the price level are not treated as bygones and must be offset. Specifically, a price-level target requires that an
overshoot of the target must be reversed and this might require quite contractionary monetary policy and, with sticky prices, this could lead to a
sharp downturn to the real economy in the short run. Indeed, if the overshoot is large enough, returning to the target might require a deflation, which
could promote financial instability and be quite harmful to the economy.
Although the models with a forward-looking price setting do not find
Two Decades of Japanese Monetary Policy and the Deflation Problem
171
that this feature of a price-level target increases output variability, they do
not focus on the fact that a price-level target may lead to more frequent
episodes of deflation, which leads to the problems discussed in section 4.2:
deflation can exacerbate financial instability or can make monetary policy
more difficult to conduct because interest rates cannot go below zero.
These costs of deflation tend to make us more skeptical about theoretical
results that indicate that price-level targets are able to reduce output variability when inflation is positive. Indeed, price-level targets which lead to
more episodes of deflation may be more dangerous than their proponents
have realized.
In addition, a price-level target may be more difficult to explain to the
public because it is a moving target, in contrast to an inflation target, which
is not. Because increased transparency and accountability is a highly desirable attribute for the conduct of monetary policy, this is an important
advantage for an inflation target.
Another problem for a price-level target that has received little attention
in the literature is the presence of measurement error in inflation. Most research on measurement error takes the view that it is inflation that is measured with error rather than the price level and this was the approach taken
by the Boskin Commission.42 This implies that the measurement error in
the price level is I(1), that a price-level target results in growing uncertainty
about the true price level as the forecast horizon grows. Thus many of the
arguments that a price-level target results in lower long-run uncertainty
about the true price level may be overstated.
The conflicting arguments above indicate that whether price-level rather
than inflation targets would produce better outcomes when inflation is positive is an open question. Given this uncertainty about the benefits of pricelevel targeting, it is not surprising that no central bank has decided to target the price level in recent years.43 However, the arguments made here for
preferring an inflation target over a price-level target do not rule out hybrid
policies, which combine features of an inflation and a price-level target and
so might provide the best of both worlds.
An inflation target could be announced with a commitment to some error correction in which target misses will be offset to some extent in the future. Recent research shows that an inflation target with a small amount of
error correction can substantially reduce the uncertainty about the price
level in the long run, but still generate very few episodes of deflation (e.g.,
Black, Macklem, and Rose 1998; King 1999; Battini and Yates 1999). Furthermore, by putting a small weight on the price-level error-correction term,
the trade-off between output and inflation fluctuations can be improved
42. See Boskin et al. (1996), Moulton (1996), and Shapiro and Wilcox (1996), for example.
43. However, a price-level target was used in the 1930s in Sweden (Berg and Jonung 1999).
172
Takatoshi Ito and Frederic S. Mishkin
(e.g., see also Williams 1999; Smets 2000; Gaspar and Smets 2000; McLean
and Pioro 2000). Evaluating hybrid policies of this type is likely to be a major focus of future research.
One issue that would have to be addressed if such a hybrid policy was
adopted is how it could be explained to the public. As is emphasized in Bernanke and Mishkin (1997), Mishkin (1999a), and Bernanke, et al. (1999),
critical to the success of inflation targeting is that it provides a vehicle for
more effective communication with the public. The public will clearly
not understand the technical jargon of error-correction models. However,
some form of an error-correction feature of an inflation-targeting regime
could be communicated by not only announcing an intermediate-term inflation target, but also by indicating that there is a target for the average
inflation rate over a longer period, say five years.
4.7 Deflation Cures
Once an economy begins to experience a deflation, it encounters an additional set of problems that alter the issues that confront monetary policy.
First is that the economy may be in a deflation trap in which monetary policy operating through short-term interest rates is powerless to extricate the
economy from the deflation because the policy interest rate cannot be
driven below the zero lower bound, which leaves the real interest rate too
high to stimulate recovery. Second, the central bank may have a severe
credibility problem in which the markets and public are unconvinced that
monetary policy can be committed to future expansion that would return
the economy to health. Both of these problems are exactly what we see in
Japan today. As we discussed in section 4.1, the Japanese economy is still
experiencing deflation even though short-term interest rates are at zero,
while past BOJ policies, particularly under Hayami, have suggested to the
public that once there is a glimmer of recovery, the monetary authorities
are likely to raise interest rates and tighten monetary policy.
Given these problems, what can be done to get the economy out of the
deflationary spiral? We will discuss two key elements of strategies to cure
deflation: (a) management of expectations through adoption of a pricelevel target, and (b) nonconventional policies that employ central bank
purchases of other assets besides short-term bonds.
4.7.1 Price-Level Targets
According to traditional monetary theory, it might appear as though
monetary policy cannot be effective in escaping the deflation trap because
there is no way to drive the standard interest-rate instrument below zero.
However, recent literature (Krugman 1998; Eggertsson and Woodford
2003; Auerbach and Obstfeld 2003; Svensson 2003) suggests that there is a
solution to this problem: management of expectations. If the central bank
Two Decades of Japanese Monetary Policy and the Deflation Problem
173
can convince the markets and the public that there will be higher inflation
in the future, then even with the interest rate at a floor of zero, the real interest rate will fall and this will stimulate aggregate demand through the
usual channels (Mishkin 1996). But how is the central bank to do this?
Once an economy has entered a prolonged deflation as it has in Japan,
lowering the real interest rate to stimulate the economy requires a substantial increase in expected inflation. This is why Krugman (1998) made
the radical suggestion for the BOJ to adopt an inflation target of 4 percent
for a fifteen-year period. However, a high inflation target, as suggested by
Krugman, is unlikely to be credible for two reasons. First, a commitment
to a high inflation target may not be credible because it is too much at variance with a goal of price stability. As documented in Bernanke, Laubach,
Mishkin, and Posen (1999), no inflation-targeting central bank in an industrialized country has chosen a medium-term inflation target above 3
percent. Indeed, we suspect that the Krugman proposal may have increased the BOJ’s resistance to inflation targeting because this level of inflation was well above what officials in the bank believed was consistent
with price stability. Furthermore, once the economy has emerged from a
deflationary spiral and starts to recover, the central bank will be tempted
to renege on its commitment to a high inflation target because it would like
the economy to return to an inflation rate consistent with price stability.
Thus as pointed out by Eggertsson (2003), a central bank in a deflationary
environment is subject to a time-inconsistency problem: it cannot credibly
commit to “being irresponsible” and so continue to shoot for high inflation. The result of the time-inconsistency problem is that the markets
would not be convinced the inflation would remain high, inflation expectations would not be sufficiently high to lower real rates sufficiently to stimulate the economy out of the deflation trap.
Another problem with an inflation target is that it is not “historydependent” because it is purely forward-looking (Woodford 2000, 2003).
An inflation target is not adjusted depending on the past outcome of inflation, and, as Eggertsson and Woodford (2003) have shown, will not be
effective in extricating an economy from a deflation trap. When the interest rate has hit a floor of zero, a deflationary shock, which lowers the price
level and puts the economy even farther below its potential, requires an
even higher expected inflation in order for the real interest rate to be lowered and be even more stimulative. A price-level target does exactly this:
with a price-level target, the same price-level target implies that inflation
will be expected to be higher, and this produces exactly the right response
of a lower real interest rate and more stimulative monetary policy.
The theoretical argument for a price-level target when an economy is in
a deflationary environment is thus quite strong. But there is a further reason for adoption of a price-level target when an economy has experienced
a prolonged period of deflation along with a severe balance-sheet problem
174
Takatoshi Ito and Frederic S. Mishkin
that prevents the financial system from working properly as in Japan (e.g.,
Posen 1998; Mishkin 1998; Hoshi and Kasyhap 2005). In Japan, nonperforming loans have weakened bank balance sheets, and the lack of capital
has meant that banks have been forced to cut back on lending, particularly
for new investment. The result is that the financial system is unable to allocate capital to productive investment opportunities, and this is a key element in the stagnation in Japan. The deflation has also weakened corporate
balance sheets, who have found their debt increase in value in real terms
while their assets have not (the debt-deflation phenomenon described by
Irving Fisher [1933]). The loss of net worth implies that even firms with
good investment opportunities may then not be able to get funds at favorable rates because the firm is more likely to engage in risky (moral hazard)
behavior because there is less at stake in the firm (Mishkin 1997). Thus
restoring both financial and nonfinancial balance sheets is crucial to helping an economy like Japan’s to achieve a more efficient allocation of capital that will restore it to health.
A price-level target that would get the price level to what it would have
been if the economy had not experienced deflation is an important way to
help restore balance sheets. A higher price level would lead to lower real indebtedness of firms and would thereby increase their net worth, making it
more attractive to lend to them if they have productive investment opportunities. The improvement in firms’ balance sheets would also help reduce
nonperforming loans which would have a positive knock-on effect on bank
balance sheets, thus making it easier for them to lend.
Furthermore, after a prolonged period of deflation, an economy may
need to undergo substantial restructuring if it is to return to health. Both
the BOJ and commentators on the Japanese economy have stressed the
need for restructuring of the Japanese economy.44 Indeed, the BOJ has continually argued that the economy cannot recover without restructuring
and has worried that expansionary monetary policy was seen as an alternative to the needed restructuring and thus may be counterproductive.
(This rhetoric seems to have stopped under Governor Fukui’s leadership
after March 2003.) Closing down inefficient firms and financial institutions may be exactly what the economy needs in the long run, but in the
short-run it might lead to severe dislocations and unemployment. Indeed,
this is probably why there has been so much resistance to the restructuring
process on the part of Japanese politicians. Here is where a price-level target to raise the price level comes in. As we have seen, a higher price level
would help restore financial and nonfinancial balance sheets and would
help the financial system to start working again to allocate capital, which
is critical to a restructuring process. Also, to the extent that a commitment
44. See, for example, Yamaguchi (1999).
Two Decades of Japanese Monetary Policy and the Deflation Problem
175
to a higher price level by the monetary authorities helps raise aggregate demand, this would help cushion the short-term negative effects of the restructuring process. A price-level target that encourages more expansionary monetary policy is thus more sensibly viewed as a complement to
restructuring rather than an impediment.
The analysis above suggests that a price-level target has many advantages when an economy is already experiencing deflation. Also in this case,
the criticism that a price level might lead to an overshoot of the target that
must be reversed, which could lead to deflation and an economic contraction, is no longer valid. When an economy is in a deflation trap and is far
from the appropriate price-level target, the price level is necessarily lower
than the target and so it promotes higher expected inflation which lowers
real interest rates, and this then works in exactly the right direction to get
the economy back on track. A price-level target thus dominates an inflation target in a deflationary environment.
Note that since October 1997, the CPI excluding fresh food has fallen by
3.7 percent to the present, while annual averages of the CPI has fallen by
2.5 percent between 1998 and 2003. This certainly understates the amount
of deflation because, as is well known, measured inflation is likely to be an
upward-biased measure of true inflation.45 Most estimates of measurement
error in CPI inflation in industrialized countries is around 1 percent and a
similar finding has been found for Japan (Shiratsuka 1999). Hence we regard an annual increase in measured CPI at or around 1 percent as absolute price stability. So this would suggest that a target for the CPI would
be 11.7 percent over current (March 2005) levels.46 However, because the
price-level target is a moving target it would continue to rise at the 1 percent rate and so the cumulative price increase when the target is reached
would necessarily be higher in the future.
Let us illustrate our point for a hypothetical price target. Suppose that
the price-level target was reached five years in the future, by March 2010.
The cumulative increase of the CPI over the five years would need to be
17.4 percent (which includes the cumulative increase in the target over five
years of 1 percent a year, 5.1 percent).47 If this target is credible, this would
mean that expected inflation would be 3.3 percent over the five years, and
so seven with a nominal interest rate of zero, the real interest rate would fall
to –3.3 percent, which would be highly stimulative, exactly along the lines
that Eggertsson and Woodford (2003) suggest would be appropriate.
45. The CPI excluding fresh food was 101.1 in October 1997, which turned out to be a peak.
In February 2004, the index became 97.5 (3.5 percent lower than six years ago). The annual
average of 1998 was 100.4, while it was 98.0 in 2003. The CPIexFood level of 2003 was less
than the peak by 2.4 percent.
46. 1.077 1.037 1.117, for example, an 11.7 percent increase.
47. 1.117 1.051 1.174, for example, a 17.4 percent increase.
176
Takatoshi Ito and Frederic S. Mishkin
But what should be done once the price-level target is achieved? One
strand of the literature suggests that it would be optimal to continue with
the price-level target. In models with a high degree of forward-looking behavior (e.g., Svensson 1999; Woodford 1999, 2003; Svensson and Woodford 2003; Clarida, Gali, and Gertler 1999; Dittmar, Gavin, and Kydland
1999, 2000; Dittmar and Gavin 2000; Vestin 2000; Eggertsson and Woodford 2003) a price-level target produces less output variance than an inflation target. Prices will have a long-run anchor. However, empirical evidence
(for example, Fuhrer 1997) does not clearly support forward-looking expectations formation, and models with forward-looking behavior have
counterintuitive properties that seem to be inconsistent with inflation dynamics (Estrella and Fuhrer 1998).
The other strand recommends that inflation targeting replace the pricelevel targeting once the price-level target is achieved. One reason, as argued
by Fischer (1994), is that output variability will be less in inflation targeting in a conventional model, as opposed to a heavily forward-looking
mode. A price-level target requires that an overshoot of the target must be
reversed, and this might require contractionary monetary policy which,
with sticky prices, could lead to a sharp downturn in the real economy. Ben
Bernanke (2003) seems to have advanced this position, although he is
somewhat agnostic about the switch. Another reason an inflation target
may be more desirable after the price-level target is achieved is that it is a
little easier to explain to the public, because it is not a moving target. Increased transparency and accountability is a highly desirable attribute for
the conduct of monetary policy.
4.7.2 Nonconventional Monetary Policy
Critics of inflation targeting (Friedman 2003) have argued that the concept of “managing expectations” is problematic. Why would announcing
an inflation or a price-level target pin down expectations? Aren’t actions
more important than words? Words by themselves are not enough, but neither are actions. This argues for the use of words plus actions in the conduct of monetary policy.
This raises the issue of what actions will actually influence the economy
and help make a price-level or inflation target credible, particularly when
the policy interest rate has hit a floor of zero? Once the short-term, policy
interest rate is at the floor of zero, it clearly cannot be driven lower. Thus
the conventional monetary-policy tool of manipulating the short-term,
policy interest rate is no longer an option. Is the central bank powerless?
What nonconventional policy measures can it take to affect the economy
and thereby achieve its price-level or inflation target? We look at four types
of measures below: (a) quantitative easing, (b) openmarket operations in
long-term bonds, (c) foreign exchange rate intervention, and (d) open market purchases of private, real assets.
Two Decades of Japanese Monetary Policy and the Deflation Problem
Fig. 4.6
177
M2 CD and monetary base, 1995:1–2004:1
Quantitative Easing
The nonconventional monetary policy tried by the BOJ has been the socalled “quantitative easing.” This involves an expansion of the monetary
base, even when the policy interest rate cannot be driven any lower, either
through open market operations on short-term government debt, outright
purchase of long-term bonds (or equities), or through unsterilized purchases of foreign currency. The BOJ has been conducting such a policy
since March 2001, and more aggressively since December 2001.
Figure 4.6 shows growth rates of monetary base (MB) and the money
supply (M2 CD, hereafter simply M2). MB had indeed expanded
quickly from the end of 2001, but with little impact on M2. How to explain
the deviation between MB and M2 is a challenge, and another is whether
an expansion of MB without an expansion of M2 has positive impacts on
the economy. The monetary base includes the amount of current account
at the BOJ, the amount of excess liquidity in the system. In normal times,
excess reserves would be unlikely to help stimulate the economy. However,
an expansion of the monetary base might be beneficial even if it does not
produce a significant increase in M2 when the interest rate is zero. First,
ample liquidity in the system may help avoid a potential financial crisis that
was a concern in 2002–2003. Second, liquidity may encourage financial institutions to take more risks in portfolio management, in particular taking
positions in long-term bonds, equities, and foreign bonds, any of which
would contribute to stimulating the economy indirectly. The economic recovery in 2003 may be partly due to ample liquidity in the system.
However, the data do not look favorable to this approach. The monetary
178
Takatoshi Ito and Frederic S. Mishkin
base has increased by 20–40 percent from 2002 to 2003 and yet deflation
has not stopped. One problem with coming to this conclusion based on the
evidence from Japan is that, as we have seen in the earlier section of this
chapter, the BOJ has created market expectations that even when it pursued expansionary monetary policy for a time, it would soon reverse it.
Then it is no surprise that quantitative easing would not work.
In addition, there are good theoretical reasons why quantitative easing
might be ineffective. The conventional liquidity-trap analysis suggests that
when the short-term interest rate hits a floor of zero, short-term bonds become a perfect substitute for money and so expanding the monetary base
will have no effect on the economy. Eggertsson and Woodford (2003) show
that this result can even hold if short-term bonds and money do not become perfect substitutes, although this conclusion still is based on the specific assumptions of their model. However, as they emphasize, quantitative
easing might help stimulate the economy if it provided a signal that the
monetary base would be higher than it otherwise would be once the deflation is over. This is the position taken by Auerbach and Obstfeld (2003).
Given the theoretical arguments against its being effective and the fact
that quantitative easing has not worked to stimulate the economy and stop
deflation in Japan, there is clearly a strong case that the BOJ needs to look
at other approaches to conducting monetary policy.
Open Market Operations in Long-Term Bonds
Alternative nonconventional monetary policies involve the monetary
authorities in conducting open market operations in other assets besides
short-term bonds. The most conventional of these is a shift toward central
bank purchases of long-term rather than short-term bonds. Since longterm interest rates are more likely to figure in household and business decisions about spending, it seems that open market purchase of these bonds
might succeed in lowering long-term interest rates, thereby stimulating the
economy. However, in order for purchase of long-term bonds to work there
would have to be significant portfolio-balance effects, so that a shift in the
supply of long-term versus short-term government debt in the hands of the
public as a result of the open market purchases would affect risk (term)
premiums and so result in a fall in long-term rates. However, the evidence
that risk (term) premiums can be affected by changing the supply of longterm bonds relative to short-term bonds in the hands of the public is, unfortunately, far from clear. One episode in which this was tried was the socalled “Operation Twist” in the United States in the early 1960s and it has
generally been viewed as a failure with only a very small effect—if any—
on the relative interest rates of long versus short-term bonds (see Meulendyke [1998] for a summary of the literature).
Bernanke (2002) has suggested that the apparent failure of “Operation
Twist” does not mean that the central bank could not drive long-term bond
Two Decades of Japanese Monetary Policy and the Deflation Problem
179
rates down as long as the central bank announced that it would peg interest rates on long-term bonds at a very low interest rate (possibly zero) and
stood ready to purchase any amounts of these bonds at this low rate. This
peg could certainly work because the commitment is easily verifiable since
the price and interest rates on long-term bonds are immediately known.
However, this could require the central bank to purchase the entire stock
of long-term bonds which it might not be fully comfortable about doing.
Clearly another way for the central bank to lower long-term bond rates
(Orphanides and Wieland 2000) is to convince the markets that it will continue to pursue a zero interest rate policy (ZIRP) for a considerable time
even after the deflation is over. Then, as is suggested by the expectations hypothesis of the term structure, because long-term bond rates are an average of the expected future short-term rates, long-term interest rates would
necessarily fall. Indeed, this strategy is complimentary to Bernanke’s
(2002) because it is a way of committing to more expansionary policy in the
future even after the economy has bounced back.
The BOJ’s announcements about clarifying the conditions of exit from
the ZIRP have some elements of this strategy. The BOJ has effectively announced that it will not reverse the ZIRP policy until there is clear-cut evidence that the deflation is over and that it is unlikely to recur in the future.
However, this is a far weaker commitment than the strategy above suggests.
It requires a commitment to stay with ZIRP not only until the deflation is
clearly over, but until there is a prospect of achieving the price-level target
described above in which the CPI would have to rise substantially to get to
the target.48 There is still the problem that an announcement of this type
might not be believed by the markets because of the past behavior of the central bank, and this is clearly a problem for the BOJ because of the policies
under Governor Hayami where the ZIRP was reversed as soon as the economy began to recover. However, this is where the purchase of long-term
bonds might help. The central bank could buy substantial amounts of these
long-term bonds as a signal of its confidence that their price will remain high
because ZIRP will be continued well after the deflation is over. Buying longterm bonds would also provide incentives for the central bank to stick with
the ZIRP policy after the deflation is over because premature abandonment
of ZIRP would lead to losses on the long-term bonds that it has bought.
Foreign Exchange Intervention
Depreciation of the currency provides an additional way of exiting from
a deflation trap. A fall in the value of the domestic currency makes imports
more expensive and exports cheaper. The result is expenditure switching in
48. In order not to overshoot the target, ZIRP would have to be removed a little while before the target is reached, but for all practical purposes, this would be a commitment to keep
ZIRP for a substantial period after the deflation is over.
180
Takatoshi Ito and Frederic S. Mishkin
which exports rise and imports fall, thereby increasing the demand for domestically produced goods, which stimulates aggregate demand. Intervention in the foreign exchange market, the selling of yen and purchase of foreign currency, has thus been suggested as a powerful way of getting the
Japanese economy moving again (Bernanke 2000; McCallum 2000a, b,
2002, 2003; Meltzer 2001; Orphanides and Wieland 2000; Svensson 2001,
2003). Indeed, in recent years the Ministry of Finance and the BOJ intervened in the foreign exchange market to keep the yen from appreciating,
but have not engineered a depreciation of the yen.
One problem with this transmission mechanism is that it also requires
that portfolio-balance effects are operational. The exchange rate intervention in which the purchase of foreign-denominated assets are bought with
domestic currency, thereby increasing the supply of domestic currencydenominated assets relative to foreign-denominated assets, only affects
the exchange rate if domestic and foreign assets are imperfect substitutes.
As was the case for short-term versus long-term bonds, the evidence for
portfolio-balance effects are not strong (see the survey in Sarno and Taylor 2001).
However, here is where a price-level target and the management of expectations can again come to the rescue. Svensson (2001, 2003) has advocated that, along with an announcement of a price-level target along the
lines we have described above, the government and/or the central bank (depending on who controls foreign exchange intervention) commit to an exchange rate peg that is consistent with that price-level target. This involves
a commitment to an immediate depreciation of the domestic currency,
which would then be allowed to appreciate at the rate of the foreign interest rate differential (so that the expected return on foreign and domestic assets is equalized). The peg would then be abandoned once the price-level
target has been achieved and a price-level or inflation-targeting regime
would be put into place. Committing to the peg is also a commitment to
the higher price-level target and continued expansionary monetary policy
even after the deflation is over. Thus it solves the commitment problem described above.
Since the policy calls for a substantial depreciation of the domestic currency from current levels, it would require that the government or central
bank stand ready to buy large amounts of foreign-denominated assets to
ensure that they are a good investment relative to domestic assets. This
would just mean an accumulation of international reserves, which is always
feasible. (This is in contrast to a case in which a country wants to prop up
the value of its currency and thus must sell foreign assets, thereby losing international reserves, which may run out and thus force the abandonment
of the peg.) The commitment to a peg also has the advantage that it provides incentives for the central bank and the government to stick with the
peg until the price-level target is achieved: early abandonment would lead
Two Decades of Japanese Monetary Policy and the Deflation Problem
181
to an appreciation of the domestic currency, which would result in substantial losses on the holdings of international reserves.
Although we agree with Svensson that his “foolproof way” to escape the
deflation trap would work, we do have our doubts about this strategy. Such
a strategy suffers from two difficulties. First, the country’s trading partners
would be likely to be up in arms if an exchange-rate peg of this type were
announced. We have seen strong U.S. complaints against the Chinese peg
of the yuan at, most likely, an undervalued rate, and we expect that this outcry would be even harsher if Japan adopted Svensson’s suggestion. The
outcome might be trade sanctions and a rise in protectionism that could be
disastrous for the world trading system.
A second problem is that adoption of an exchange rate peg might cause
a shift of the nominal anchor away from the price-level or inflation target
to the exchange rate. For example, as part of its inflation-targeting regime,
Israel has had an intermediate target of an exchange rate band around a
crawling peg, whose rate of crawl is set in a forward-looking manner by deriving it from the inflation target for the coming year. Even though the Bank
of Israel downplayed the exchange rate target relative to the inflation target
over time, it did slow the bank’s efforts to win support for disinflation and
lowering of the inflation targets (see Bernanke, Laubach, Mishkin, and
Posen 1999). A recent example of this problem has occurred in Hungary
(Jonas and Mishkin 2004), which has an exchange rate band as part of its
inflation-targeting regime. In January 2003, the forint appreciated to the
upper end of the band, and speculation about the revaluation of the parity
resulted in a sharp acceleration of capital inflows that forced the National
Bank of Hungary to respond by cutting interest rates by 2 percentage
points and intervening heavily in the foreign exchange market. The National Bank of Hungary is reported to have bought more than 5 billion euros, increasing international reserves by 50 percent and base money by 70
percent. (See J. P. Morgan [2003].) Even though the National Bank of
Hungary subsequently began to sterilize this huge injection of liquidity,
market participants then assumed that maintaining the exchange rate band
would have a priority over the inflation target and expected inflation in
2003 to exceed the National Bank of Hungary’s inflation target by 5 percentage points.49
A third problem with an exchange rate target is that it can induce the
wrong policy response when a country is faced with real shocks, such as a
terms-of-trade shock. Two graphic examples occurred in New Zealand and
Chile in the late 1990s. By early 1997, the Reserve Bank institutionalized
this focus by adopting as its primary indicator of monetary policy at Mon49. Analysts have interpreted this as evidence that the National Bank of Hungary is determined to maintain the currency band even at the cost of temporary higher inflation. See Jonas
and Mishkin (2004).
182
Takatoshi Ito and Frederic S. Mishkin
etary Conditions Index (MCI) similar to that developed by the Bank of
Canada. The idea behind the MCI, which is a weighted average of the exchange rate and a short-term interest rate, is that both interest rates and
exchange rates on average have offsetting impacts on inflation. When the
exchange rate falls, this usually leads to higher inflation in the future, and
so interest rates need to rise to offset the upward pressure on inflation.
However, the offsetting effects of interest rates and exchange rates on inflation depend on the nature of the shocks to the exchange rates. If the exchange rate depreciation comes from portfolio considerations, then it does
lead to higher inflation and needs to be offset by an interest rate rise. However, if the reason for the exchange rate depreciation is a real shock, such
as a negative terms-of-trade shock, which decreases the demand for a
country’s exports, then the situation is entirely different. The negative
terms-of-trade shock reduces aggregate demand and is thus likely to be deflationary. The correct interest rate response is then a decline in interest
rates, not a rise as the MCI suggests.
With the negative terms-of-trade shock in 1997, the adoption of the MCI
in 1997 led to exactly the wrong monetary-policy response to East Asian
crisis. With depreciation setting in after the crisis began in July 1997 after
the devaluation of the Thai baht, the MCI began a sharp decline, indicating that the Reserve Bank needed to raise interest rates, which it did by over
200 basis points. The result was very tight monetary policy, with the
overnight cash rate exceeding 9 percent by June of 1998. Because the depreciation was due to a substantial, negative terms-of-trade shock that decreased aggregate demand, the tightening of monetary policy, not surprisingly, lead to a severe recession and an undershoot of the inflation-target
range with actual deflation occurring in 1999.50 The Reserve Bank of New
Zealand did eventually realize its mistake and reversed course, sharply
lowering interest rates beginning in July 1998 after the economy had entered a recession, but by then it was too late.
Chile’s inflation-targeting regime also included a focus on limiting exchange rate fluctuations by having an exchange rate band with a crawling
peg that was (loosely) tied to lagged domestic inflation. This focus on the
exchange rate induced a serious policy mistake in 1998 because the central
bank was afraid it might lose credibility in the face of the financial turmoil
if it allowed the exchange rate to depreciate after what had taken place in financial markets after the East Asian crisis and the Russian meltdown. Thus
instead of easing monetary policy in the face of the negative terms-of-trade
shock, the central bank raised interest rates sharply and even narrowed its
exchange rate band. The result was that the inflation target was undershot
and the economy entered a recession for the first time in the 1990s. With
50. The terms-of-trade shock, however, was not the only negative shock the New Zealand
economy faced during that period. Its farm sector experienced a severe drought, which also
hurt the economy. Thus, a mistake in monetary policy was not the only source of the recession. Bad luck played a role too. See Drew and Orr (1999) and Brash (2000).
Two Decades of Japanese Monetary Policy and the Deflation Problem
183
this outcome, the central bank came under strong criticism for the first time
since it had adopted its inflation-targeting regime in 1990, weakening support for the independence of the central bank and its inflation-targeting
regime. During 1999, the central bank did reverse course, easing monetary
policy by lowering interest rates and allowing the peso to decline.
The contrast of the experience of New Zealand and Chile during this period with that of Australia, another small open economy with an inflationtargeting regime, is striking. Prior to adoption of their inflation-targeting
regime in 1994, the Reserve Bank of Australia had adopted a policy of allowing the exchange rate to fluctuate without interference, particularly if
the source of the exchange rate change was a real shock, like a terms-oftrade shock. Thus when faced with the devaluation in Thailand in July
1997, the Reserve Bank recognized that it would face a substantial negative
terms-of-trade shock because of the large component of its foreign trade
conducted with the Asian region and that it should not fight the depreciation of the Australian dollar that would inevitably result.51 Thus in contrast
to New Zealand, it immediately lowered the overnight cash rate by 50 basis points to 5 percent and kept it near this level until the end of 1998, when
it was lowered again by another 25 basis points.
A more subtle approach to exchange rate intervention can avoid some of
the problems of an exchange rate peg. Intervention in the foreign exchange
market to depreciate the domestic currency could be an important element
of nonconventional monetary policy of raising price-level expectations,
without announcing a precise exchange rate target. Instead the central
bank and the government could emphasize that exchange rate interventions, along with other measures, are being conducted as a method of pursuing expansionary monetary policy and to achieve a higher price level and
a stronger economy. These interventions would then be unsterilized in order to make clear that their primary purpose is to produce expansionary
monetary policy that raises the price level and is not focused on a target
level of the exchange rate.52 The communication strategy would also be
helped by having the government and the central bank emphasize that the
51. See MacFarlane (1999) and Stevens (1999).
52. Under the Hayami regime, the BOJ resisted the suggestion that interventions be unsterilized. Since interventions are decided and conducted by the Ministry of Finance, making
interventions unsterilized was seen as a dictation of monetary policy by the Ministry of Finance—a violation to independence. In 2003, under the Fukui regime, interventions became
more frequent on the part of the Ministry of Finance, and quantitative easing was accelerated
on the part of the BOJ. From January 2003 to December 2003, about 15 trillion yen of interventions increased the yen in the market in exchange for an increase in inventory of foreign
currencies, while the ceiling of the BOJ current account target was raised by 12 trillion yen.
Deputy Governor Iwata in his reply to a question in the press conference on October 1, 2003
acknowledged that these two actions, ex post, were equivalent to unsterilized interventions,
although “it must be a coincident.” This is a much more nuanced statement than a typical reaction during the Hayami regime (press interview, October 1, 2003, available in Japanese text
through the BOJ homepage at [http://www.boj.or.jp/press/03/kk0310a.htm], translated by
one of the authors of this paper).
184
Takatoshi Ito and Frederic S. Mishkin
exchange rate interventions to escape from the deflation trap would eventually help encourage purchases of foreign goods and would eventually be
highly beneficial for the country’s trading partners.
At the zero interest rate, the differences between sterilized and unsterilized intervention, namely the interest rate channel, disappear. However,
even at the zero interest rate, we believe that the unsterilized intervention is
more stimulative than sterilized intervention, primarily due to a signaling
effect. Unsterilized intervention provides one more instrument to achieve
quantitative easing, and conducting unsterilized intervention will make the
central bank look more willing to commit to continuing ZIRP in the future.
Open Market Purchase of Private, Real Assets
An even more radical step for monetary authorities would be to purchase real assets, such as stocks, corporate bonds, or real estate. Purchase
of these assets would raise their prices directly and would lead to expansion
in aggregate demand though a number of channels of monetary transmission (Mishkin 1996). Purchase of real assets would also directly help restore balance sheets in the economy and help get the financial system working again, which we have seen is crucial to recovery if the country finds
itself in a situation like Japan’s.
However, central bank purchase of these assets is not without problems.
Government purchase of private assets can be highly politicized. Which assets should the central bank buy? Different elements in the private sector
would lobby for purchase of the assets that would make them profits. Some
of this problem could be mitigated by the central bank buying broad-based
bundles of assets or market indexes so that specific private firms do not
benefit over others. However, there is still the question of how much real estate should be bought versus stocks, or how much corporate bonds versus
equities. Decisions on what to buy would have important distributional
consequences, which would put the central bank under intense political
pressure. Not only might this result in distortionary decisions, but it could
politicize the central bank and interfere with the independence that this institution has worked so hard to get.
Another problem with central bank purchase of private assets is that it
involves the government in ownership of the private sector. The trend in recent years has been toward privatization because it is believed that the
private sector has better incentives to produce efficiently than does the government sector. Having substantial purchases of private assets by the central bank, which after all is a government entity, goes against this trend.
Maybe the problems of central bank ownership of private assets can be
minimized by announcing that the central bank will have no involvement
in running of the companies or real estate that it has taken a position in,
but political pressures may make this hard to do.
If central bank purchases of private, real assets are sizeable, there could
Two Decades of Japanese Monetary Policy and the Deflation Problem
185
be adverse consequences both for the central bank and the economy. However, if nothing else worked, then this more radical step might be necessary
as a way of stimulating the economy and achieving a higher price level.
Our discussion here has indicated that none of the nonconventional
monetary-policy strategies are without their problems. There is thus an argument for what might be crudely described as a “kitchen sink” or “throw
it against the wall and see if it sticks” approach. Because it would not be
clear how well the different approaches would work, some or all of them
could be tried to see which ones work best. One concern might be that the
uncertainty about the impact of the different approaches might make it
harder to be sure of what the outcome of using them might be. One outcome would be paralysis and then not to try any of them.
There are two responses to these concerns. The first is that having a
clear-cut price-level/inflation target to pin down expectations can make it
highly likely that less conventional tools of monetary policy can achieve
the goal of price stability and that inflation would not spin out of control.
In recent years we have seen major successes in the ability of monetary policy to control inflation in many industrialized countries. We would argue
that this is not because central banks have become so much more knowledgeable about the transmission mechanisms of monetary policy. What
has changed in recent years is that central banks in industrialized countries
have been able to put much stronger nominal anchors in place. The result
is greatly improved performance on both the inflation and output fronts.
One method has been to adopt inflation targets, as in New Zealand,
Canada, the United Kingdom, Sweden, and Australia, and to some extent
in the European Monetary Union.53 Alternatively, a strong nominal anchor can be put into place without a formal inflation target through direct
communication with the public about the commitment to price stability
and actions that are consistent with it. This is the strategy pursued by the
Federal Reserve, which has as strong a nominal anchor as inflationtargeting, central banks although it is embodied in an individual, Alan
Greenspan (Mishkin 2000). Adopting a price-level target and then possibly moving to an inflation target would go a long way to ensuring an escape
from the deflation trap, while making it highly unlikely that inflation would
spin out of control thereafter.
4.8 Concluding Remarks
This chapter reviews the experience of Japanese monetary policy over
the last two decades with an emphasis on the experience of deflation from
53. The European Central Bank does not officially call their monetary policy strategy “inflation targeting,” but it is pretty close: there is a strong commitment to price stability and an
explicit inflation goal of “less than but close to 2%” has been announced.
186
Takatoshi Ito and Frederic S. Mishkin
the mid-1990s. The cost of deflation is quite high, and prolonged deflation
makes getting out of it more difficult. A key element in escaping deflation
is the management of expectations and we have seen that price-level and inflation targeting attempt to achieve exactly this. Also, because the credibility of price-level and inflation targets require actions, nonconventional
policy measures become relevant when prices are declining and the zero
lower bound on interest rates means that the overnight interest rate can no
longer be used as the instrument of monetary policy.
We are quite critical of the conduct of the BOJ monetary policy from
1998 to 2003. The Bank of Japan’s rhetoric was not helpful in fighting deflation, and the interest rate hike in August 2000 amid deflation was a serious mistake. Although rhetoric has improved since 2003 under the new
Governor Fukui, more is needed to get out of deflation completely. We surveyed the literature on cost of deflation, the optimal level of inflation, and
relative merits of price-level versus inflation targets. A key to curing deflation is management of expectations, and here a history-dependent policy
involving a price-level target can help. However, because actions speak
louder than words, management of expectations also involves nonconventional monetary policies. Admittedly, there is uncertainty about how these
policies would work, but a combination of them can help the Japanese
economy escape its deflationary trap.
References
Ahearne, Alan, Joseph Gagnon, Jane Haltmaier, Steve Kamin, and others. 2002.
Preventing deflation: Lessons from Japan’s experiences in the 1990s. Board of
Governors, International Finance. Discussion Paper No. 729, June. Washington, DC: Board of Governors.
Akerlof, George, William Dickens, and George Perry. 1996. The macroeconomics
of low inflation. Brookings Papers on Economic Activity, Issue no. 1:1–59.
Atkeson, Andrew, and Patrick J. Kehoe. 2004. Deflation and depression: Is there an
empirical link? NBER Working Paper no. 10268. Cambridge, MA: National Bureau of Economic Research, January.
Auerbach, Alan J., and Maurice Obstfeld. 2003. The case for open market purchases in a liquidity trap. NBER Working Paper no. 9814. Cambridge, MA: National Bureau of Economic Research, July.
Bank of Japan. 2000. On Price Stability. [http://www.boj.or.jp/en/seisaku/05/
seisak_f.htm]
Battini, N., and T. Yates. 1999. Hybrid inflation and price level targeting. Bank of
England. Mimeo.
Beebe, Jack H. 2001. Comments on Oda, Nobuyuki and Okina, Kunio, “Further
monetary easing policies under the non-negativity constraints of nominal interest rates: summary of the discussion based on Japan’s experience.” Monetary and
Economic Studies 19 (S-1): 365–8. Bank of Japan.
Berg, Claes, and Lars Jonung. 1999. Pioneering price level targeting: The Swedish
experience 1931–1937. Journal of Monetary Economics 43:525–51.
Two Decades of Japanese Monetary Policy and the Deflation Problem
187
Bernanke, Ben S. 1983. Nonmonetary effects of the financial crisis in the propagation of the Great Depression. American Economic Review 73:257–76.
———. 2000. Japanese monetary policy: A case of self-induced paralysis? In Japan’s
financial crisis and its parallels to U.S. experience, ed. Ryoichi Mikitani and Adam
S. Posen, 149–66. Washington, DC: Institute for International Economics.
———. 2002. Deflation: Making sure “it” doesn’t happen here. Remarks before the
National Economists’ Club, 21 November, Washington, DC.
———. 2003. Some thoughts on monetary policy in Japan. Speech before the
Japan Society of Monetary Economics, 31 May, Tokyo, Japan. Available at
[http://www.federalreserve.gov/boarddocs/speeches/2003/20030531/default.htm]
Bernanke, Ben S., and Mark Gertler. 1999. Monetary policy and asset price volatility. Federal Reserve Bank of Kansas City, Economic Review 4th Quarter: 17–51.
[Also reprinted in Federal Reserve Bank of Kansas City, New Challenges for
Monetary Policy: 77–128.] Kansas City, MO.
Bernanke, Ben S., Thomas Laubach, Frederic S. Mishkin, and Adam S. Posen.
1999. Inflation targeting: Lessons from the international experience. Princeton,
NJ: Princeton University Press.
Bernanke, Ben S., and Frederic S. Mishkin. 1997. Inflation targeting: A new framework for monetary policy? Journal of Economic Perspectives 11 (2): 97–116.
Black, Richard, Tiff Macklem, and David Rose. 1998. On policy rules for price stability. Price stability, inflation targets and monetary policy. Proceedings of a Conference held by Bank of Canada, May 1997, Ottawa, Canada: 411–61.
Blinder, Alan S. 1998. Central banking in theory and practice. Cambridge, MA:
MIT Press.
Brash, Donald T. 2000. Inflation targeting in New Zealand, 1988–2000. Speech to
the Trans-Tasman Business Cycle, Melbourne, February 9.
Boskin, Michael J., Ellen R. Dulberger, Robert J. Gordon, Zvi Griliches, and Dale
W. Jorgenson. 1996. Toward a more accurate measure of the cost of living. Final
Report to the Senate Finance Committee, December. Washington, DC: U.S.
Government. Available at [http://www.ssa.gov/history/reports/boskinrpt.html]
Card, D. and D. Hyslop. 1997. Does inflation “grease the wheels of the labor market”? In Reducing inflation: Motivation and strategy, ed. C. D. Romer and H. D.
Romer, 71–114. Chicago: University of Chicago Press.
Cargill, Thomas, Michael Hutchison, and Takatoshi Ito. 1997. The political economy of Japanese monetary policy. Cambridge, MA: MIT Press.
———. 2000. Financial policy and central banking in Japan. Cambridge, MA: MIT
Press.
Cecchetti, Stephen G., Hans Genberg, John Lipsky, and Sushil Wadhwani. 2000.
Asset prices and central bank policy, Geneva Reports on the World Economy no.
2. Geneva: International Center for Monetary and Banking Studies.
Clarida, Richard J., Jordi Gali, and Mark Gertler. 1998. Monetary rules in practice: Some international evidence. European Economic Review 42:1033–67.
———. 1999. The science of monetary policy: A new-Keynesian perspective. Journal of Economic Literature 37:1661–707.
Clouse, James, Dale Henderson, Athanasios Orphanides, David Small, and Peter
Tinsley. 2000. Monetary policy when the nominal short-term interest rate is zero.
Board of Governors of the Federal Reserve System, International Finance. Discussion Paper no. 729. Washington, DC: Board of Governors of the Federal Reserve System.
Crawford, A., and A. Harrison. 2000. Testing for downward rigidity in nominal
wage rates. In Price stability, inflation targets and monetary policy, 179–218. Proceedings of a conference held by the Bank of Canada, May 1997 Ottawa: Bank
of Canada.
188
Takatoshi Ito and Frederic S. Mishkin
Dittmar, Robert and William Gavin. 2000. What do New-Keynesian Phillips curves
imply for price-level targeting? Federal Reserve Bank of St. Louis Review 82 (2):
21–30.
Ditmar, Robert, William T. Gavin, and Finn E. Kydland. 1999. The inflationoutput variability tradeoff and price level targets. Federal Reserve Bank of St.
Louis Review: 23–31.
Drew, Aaron, and Adrian Orr. 1999. The reserve bank’s role in the recent business
cycle: Actions and evolution. Reserve Bank of New Zealand Bulletin, 62 (1): 5–24.
Eggertsson, Gauti. 2003. Fighting deflation at zero nominal interest rates: Committing to being irresponsible. IMF Working Paper no. WP/03/64. March 1.
Eggertsson, Gauti B., and Michael Woodford. 2003. Optimal monetary policy in a
liquidity trap. NBER Working Paper no. 9968. Cambridge, MA: National Bureau of Economic Research, September.
Erceg, Christopher, and Andrew Levin. 2001. Imperfect credibility and inflation
persistence. Finance and Economics Discussion Series no. 2001-45. Washington, DC: Federal Reserve Board, October.
Estrella, Arturo, and Jeffrey Fuhrer. 1998. Dynamic inconsistencies: Counterfactual implications of a class of rational expectations models. Federal Reserve
Bank of Boston. Working Paper no. 98/05, July. Boston, MA.
Fares, Jean, and Thomas Lemieux. 2000. Downward nominal-wage rigidity: A critical assessment and some new evidence from Canada. In Price stability and the
long-run target for monetary policy, 3–42. Proceedings of a seminar held by the
Bank of Canada, June 2000. Ottawa: Bank of Canada.
Feldstein, Martin. 1997. Capital income taxes and the benefits of price stability.
NBER Working Paper no. 6200. Cambridge, MA: National Bureau of Economic
Research, September.
Fischer, Irving. 1933. The debt-deflation theory of great depressions. Econometrica
1:337–57.
Fischer, Stanley. 1994. Modern central banking. In The future of Central Banking,
ed. Forest Capie, Charles Goodhart, Stanley Fischer, and Norbert Schnadt, 262–
308. Cambridge, UK: Cambridge University Press.
Friedman, Benjamin. 2003. Discussion of “inflation targeting”: A critical view. presented at the Federal Reserve Bank of St. Louis Annual Conference, “Inflation
Targeting: Prospects and Problems,” October 16–17. St. Louis, MO.
Fuhrer, Jeffrey C. 1997. The (un)importance of forward-looking behavior in price
specifications. Journal of Money, Credit, and Banking 29 (3): 338–50.
Fuhrer, Jeffrey C., and Brian Madigan. 1997. Towards a compact, empiricallyverified rational expectations model for monetary policy analysis. CarnegieRochester Conference Series on Public Policy 47:197–230.
Fujiki, Hiroshi, Kunio Okina, and Shigenori Shiratsuka. 2001. Monetary policy
under zero interest rate: Viewpoints of central bank economists. Bank of Japan,
Monetary and Economic Studies February: 89–130.
Fukui, Toshihiko. 2003. Challenges for monetary policy in Japan. Speech at the
Spring Meeting of the Japan Society of Monetary Economics, June 1, Tokyo,
Japan.
Gaspar, Vitor, and Frank Smets. 2000. Price level stability: Some issues. European
Central Bank. Mimeo.
Goodfriend, Marvin. 2001. Financial stability, deflation, and monetary policy.
Bank of Japan, Monetary and Economic Studies 19 (S-1): 143–76.
Greenspan, Alan. 2002. Issues for monetary policy. Remarks by Chairman Alan
Greenspan before the Economic Club of New York, New York City.
———. 2004. Risk and uncertainty in monetary policy. Remarks by Chairman
Two Decades of Japanese Monetary Policy and the Deflation Problem
189
Alan Greenspan at the meeting of the American Economic Association, San
Diego, California, January 3.
Groshen, Erica L., and Mark E. Schweitzer. 1996. The effects of inflation on wage
adjustments in firm-level data: Grease or sand? Federal Reserve Bank of New
York Staff Reports, No. 9. New York: New York.
———. 1999. Identifying inflation’s grease and sand effects in the labor market. In
The costs and benefits of price stability, ed. Martin Feldstein, 273–308. Chicago:
University of Chicago Press.
Harrigan, James, and Kenneth Kuttner. 2004. Lost decade in translation: Has the
U.S. learned from Japan’s post-bubble mistakes? Presented at the Solution Conference, Tokyo, June.
Hoshi, Takeo, and Anil Kashyap. 2005. Solutions to the Japanese banking crisis:
What might work and what definitely will fail. In Reviving Japan’s Economy, ed.
Ito, Patrick, and Weinstein, 147–95. Cambridge, MA: MIT Press.
Hunter, William C., George G. Kaufman, and Michael Pomerleano. 2003. Asset
price bubbles: The implications for monetary, regulatory, and international policies. Cambridge, MA: MIT Press.
Ito, Takatoshi. 1992. The Japanese economy. Cambridge, MA: MIT Press.
———. 1999. A target for the bank of Japan. Financial Times, October 19.
———. 2000. Nihon ni okeru inflation targeting seisaku [Inflation targeting policy
in Japan]. In Zero kinri to nihon keizai [Zero interest rate and the Japanese economy], ed. Mitsuhiro Fukao and Hiroshi Yoshikawa, 77–127. Tokyo: Nihon
Keizai Shinbun Pub.
———. 2001. Inflation targeting (in Japanese). Tokyo: Nihon Keizai Shinbun Pub.
———. 2003. Looking forward on monetary and supervision policies to protect
against bubbles. In Asset price bubbles: The implications for monetary, regulatory,
and international policies, ed. W. C. Hunter, G. G. Kaufman, and M. Pomerleano, 547–52. Chicago: MIT Press.
———. 2004. Inflation targeting and Japan: Why has the bank of Japan not
adopted inflation targeting? In The future of inflation targeting, ed. Christopher
Kent and Simon Guttman, 220–67. Sydney: Reserve Bank of Australia.
Ito, Takatoshi, and Kimie Harada. 2000. Japan premium and stock prices: Two
mirrors of Japanese banking crises. NBER Working Paper no. 7997. Cambridge,
MA: National Bureau of Economic Research, November.
Ito, Takatoshi, and Tokuo Iwaisako. 1996. Explaining asset bubbles in Japan. The
Bank of Japan, Monetary and Economic Studies 14, July: 143–93.
Ito, Takatoshi, and Tomoko Hayashi. 2004. Inflation targeting in Asia. Hong Kong
Institute for Monetary Research, Occasional Paper, No. 1, March.
Iwata, Kazumasa. 2002. Deflation spiral hassei no kanousei [in Japanese, trans. Possibility of having a deflation spiral]. In R. Komiya, Kinyu Seisaku rongi no Soten
[in Japanese, trans. Issues in the monetary policy debate]. Tokyo: Nihon Keizai
Shinbun, pub.: 121–56.
Iwata, Kikuo. 2001. Deflation no keizaigaku, [in Japanese, trans. Economics of deflation]. Tokyo: Toyo Keizei Shinpo, Pub.
———. 2001. Deflation no keizaigaku [in Japanese, trans. Economics of deflation].
Tokyo: Toyo Keizai Shinpo, Pub.
———. 2002. Yosou keisei ni hataraki kakeru kinyu seisakuo. [in Japanese, trans.
Recommending monetary policy to affect inflation expectation]. In R. Komiya,
Kinyu seisaku rongi no soten [in Japanese, trans. Issues in the monetary policy debate]. Tokyo: Nihon Keizai Shinbun, pub.: 391–418.
Jinushi, Toshiki, Yoshihiro Kuroki, and Ryuzo Miyao. 2000. Monetary policy in
Japan since the late 19870s: Delayed policy actions and some explanations. In
190
Takatoshi Ito and Frederic S. Mishkin
Japan’s financial crisis and its parallels to U.S. experience, ed. Mikitani and Posen,
chapter 6. Washington, DC: Institute for International Economics.
Jonas, Jiri, and Frederic S. Mishkin. 2004. Inflation targeting in transition countries: Experience and prospects. In Inflation targeting, ed. Michael Woodford
(with Jiri Jonas). Chicago: University of Chicago Press, forthcoming.
Kamada, Koichiro. 2004. Real-time estimation of the output gap in Japan and its
usefulness for inflation forecasting and policy making. Bank of Japan. Mimeo.
Available at [http://www.e.u-tokyo.ac.jp/cirje/research/workshops/macro/macro
paper04/kamada.pdf]. April 16.
King, Mervyn. 1996. How should central banks reduce inflation? Conceptual issues. In Achieving price stability, Federal Reserve Bank of Kansas City, 53–91.
Kansas City, MO.
———. 1999. Challenges for monetary policy: New and old. In New challenges for
monetary policy, Federal Reserve Bank of Kansas City, 11–57. Kansas City, MO.
Krugman, Paul. 1998. It’s baack! Japan’s slump and the return of the liquidity trap.
Brookings Papers on Economic Activity Issue no. 2:137–87.
Kuroda, Sachiko, and Isamu Yamamoto. 2003a. Are Japanese nominal wages
downwardly rigid? (Part I): Examinations of nominal wage change distributions.
Monetary and Economic Studies 21 (2): 1–30. Tokyo: Bank of Japan.
———. 2003b. Are Japanese nominal wages downwardly rigid? (Part II): Examinations using a friction model. Bank of Japan, Monetary and Economic Studies
2 (21): 31–68.
———. 2003c. The impact of downward nominal wage rigidity on the unemployment rate: Quantitative evidence from Japan. Bank of Japan, Monetary and Economic Studies 21 (4): 57–85.
Kuttner, Kenneth N., and Adam S. Posen. 2004. The difficulty of discerning what’s
too tight: Taylor rules and Japanese monetary policy. The North American Journal of Economics and Finance 15 (1): 53–74.
Lebow, David E., Brian Saks, and Beth A. Wilson. 1999. Downward nominal wage
rigidity: Evidence from the employment cost index. Feds Working Paper no.
1999–31. Washington, DC: Board of Governors of the Federal Reserve System,
July.
Lebow, David E., David J. Stockton, and W. L. Wascher. 1995. Inflation, nominal
wage rigidity and the efficiency of the labor market. Finance and Economics Discussion Series, Working Paper no. 94-45. Board of Governors of the Federal Reserve Board, October.
Macfarlane, Ian J. 1999. Statement to parliamentary committee. Reserve Bank of
Australia Bulletin, January: 16–20.
McCallum, Bennett T. 1999. Issues in the design of monetary policy rules. In Handbook of macroeconomics, ed. John B. Taylor and Michael Woodford. North Holland, Amsterdam, forthcoming.
———. 2000a. Theoretical analysis regarding a zero lower bound on nominal interest rates. Journal of Money, Credit and Banking 32 (2): 879–904.
———. 2000b. Alternative monetary policy rules: A comparison with historical
settings for the United States, the United Kingdom, and Japan. Federal Reserve
Bank of Richmond, Economic Quarterly 86, Winter: 49–79.
———. 2002. Inflation targeting and the liquidity trap. In Inflation targeting: Design, performance, challenges, ed. N. Loayza and R. Soto, 395–437. Central Bank
of Chile.
———. 2003. Japanese monetary policy 1991–2001. Economic Quarterly 89 (1): 1–
31. Richmond, VA: Federal Reserve Bank of Richmond.
McLean, D., and H. Pioro. 2000. Price level targeting—The role of credibility. Paper
Two Decades of Japanese Monetary Policy and the Deflation Problem
191
presented at the Bank of Canada Conference, Price Stability and the Long-Run
Target for Monetary Policy, June 8–9, Ottawa, Bank of Canada, 153–200. Available at [http://www.bankofcanada.ca/en/conference/con2000/con2000.html]
McKinnon, Ronald I. and Kenichi Ohno. 1997. Resolving economic conflict between the United States and Japan. MIT Press.
Meltzer, Allan H. 1995. Monetary, credit and other transmission mechanism processes: A monetarist perspective. Journal of Economic Perspectives 9:49–72.
———. 1998. Time to print money. Financial Times, July 17.
———. 2001. Monetary transmission at low inflation: Some clues from Japan in the
1990s. Bank of Japan, Monetary and Economic Studies 19 (S-1): 13–34.
Meulendyke, Ann-Marie. 1998. U.S. monetary policy and financial markets. New
York: Federal Reserve Bank of New York.
Mishkin, Frederic S. 1978. The household balance-sheet and the Great Depression.
Journal of Economic History 38, December: 918–37.
———. 1991. Asymmetric information and financial crises: A historical perspective. In Financial markets and financial crises, ed. R. Glenn Hubbard, 69–108.
Chicago: University of Chicago Press.
———. 1996. The channels of monetary transmission: Lessons for monetary policy. Banque De France Bulletin Digest No. 27:33–44. Paris: Bank of France.
———. 1997. The causes and propagation of financial instability: Lessons for policymakers. In Maintaining financial stability in a global economy, Federal Reserve
Bank of Kansas City, 55–96. Kansas City, MO.
———. 1998. Promoting Japanese recovery. In Towards the restoration of sound banking systems in Japan—the global implications, ed. Kenichi Ishigaki and Hiroyuki
Hino, 130–61. Kobe: Kobe University Press and International Monetary Fund.
———. 1999a. International experiences with different monetary policy regimes.
Journal of Monetary Economics 43 (3): 579–606.
———. 1999b. Central banking in a democratic society: Implications for transition
countries. In Central banking, monetary policy and the implications for transition
economies, ed. Mario Blejer and Marko Skreb, 31–53, Boston: Kluwer Academic
Publishers.
———. 2000. What should central banks do? Federal Reserve Bank of St. Louis Review 82 (6): 1–13.
———. 2001. The transmission mechanism and the role of asset prices in monetary
policy. In Aspects of the transmission mechanism of monetary policy, focus on Austria 3–4/2001, ed. Edward Hochreiter, 58–71. Vienna: Öesterreichische National
Bank.
———. 2004a. Why the Fed should adopt inflation targeting. International Finance
7:1, forthcoming.
———. 2004b. Can central bank transparency go too far? Reserve Bank of Australia, The future of inflation targeting, forthcoming.
Mishkin, Frederic S., and Adam S. Posen. 1997. Inflation targeting: Lessons from
four countries. New York: Federal Reserve Bank of New York, Economic Policy
Review 3 (3): 9–110.
Mishkin, Frederic S., and Eugene N. White. 2003. U.S. stock market crashes and
their aftermath: Implications for monetary policy. In Asset price bubbles: The
implications for monetary, regulatory and international policies, ed. William B.
Hunter, George G. Kaufman, and Michael Pormerleano, 57–79. Cambridge,
MA: MIT Press.
Morgan, J. P. 2003. Emerging Europe, Middle East & Africa Weekly, January 31.
Moulton, Brett R. 1996. Bias in the consumer price index: What is the evidence.
Journal of Economic Perspectives 10 (4): 159–77.
192
Takatoshi Ito and Frederic S. Mishkin
Oda, Nobuyuki, and Kunio Okina. 2001. Further monetary easing policies under
the non-negativity constraints of nominal interest rates: Summary of the discussion based on Japan’s experience. Monetary and Economic Studies 19 (S-1): 323–
70. Tokyo: Bank of Japan.
Okina, Kunio. 1999a. Monetary policy under zero inflation: A response to criticisms
and questions regarding monetary policy. Monetary and Economic Studies
17:157–82. Tokyo: Bank of Japan, Institute for Monetary and Economic Studies.
———. 1999b. Rejoinder to comments made by professors McKinnon and
Meltzer. Monetary and Economic Studies 17:192–7. Tokyo: Bank of Japan, Institute for Monetary and Economic Studies.
Okina, Kunio, Masaaki Shirakawa, and Shigenori Shiratsuka. 2001. The asset
price bubble and monetary policy: Japan’s experience in the late 1980s and the
lessons. Background Paper for the conference: Monetary and Economic Studies
special edition, February: 395–450.
Okina, Kunio, and Shigenori Shiratsuka. 2002. Asset price bubbles, price stability,
and monetary policy: Japan’s experience. Bank of Japan, Monetary and Economic Studies October: 35–76.
———. 2004. Policy commitment and expectation formation: Japan’s experience
under zero interest rates. North American Journal of Economics and Finance
15:75–100.
Orphanides, Athanasios, and Volker Wieland. 1998. Price stability and monetary
policy effectiveness when nominal interest rates are bounded by zero. Feds
Working Paper No. 1998-35. Washington, DC: Board of Governors of the Federal Reserve System, August.
———. 2000. Efficient monetary policy design near price stability. Journal of Japanese and International Economies 14:327–56.
Poole, William. 1999. Is inflation too low. Federal Reserve Bank of St. Louis Review
81 (4): 3–10. St. Louis, MO.
Posen, Adam S. 1998. Restoring Japan’s economic growth. Washington, DC: Institute for International Economics.
Reifschneider, David, and John Williams. 2000. Three lessons for monetary policy
in a low inflation era. Journal of Money, Credit and Banking (November): 936–66.
Richards, Anthony, and Tim Robinson. 2003. Asset Prices and Monetary Policy.
Sydney: Reserve Bank of Australia.
Romer, Christine. 1992. What ended the Great Depression? Journal of Economic
History 52 (4): 757–84.
Sarno, Lucio, and Mark P. Taylor. 2001. Official intervention in the foreign exchange market: Is it effective and, if so, how does it work? Journal of Economic
Literature XXXIX (September): 839–68.
Shapiro, Matthew D., and David W. Wilcox. 1996. Mismeasurement in the consumer price index, an evaluation. In NBER Macroeconomics Annual 1996, ed.
Ben S. Bernanke and Julio J. Rotemberg, 83–154. Cambridge, MA: MIT Press.
Shiratsuka, Shigenori. 1999. Measurement errors and quality-adjustment methodology: Lessons from the Japanese CPI. Federal Reserve Bank of Chicago, Economic Perspectives, 2nd Quarter: 2–13. Chicago.
Smets, Frank. 2000. What horizon for price stability. European Central Bank
Working Paper No. 24. Frankfurt, Germany: European Central Bank.
Stevens, Glenn R. 1999. Six years of inflation targeting. Reserve Bank of Australia
Bulletin May: 46–61. Sydney: Reserve Bank of Australia.
Summers, Lawrence H. 1991. How should long-term monetary policy be determined? Journal of Money Credit and Banking 23:625–31.
Svensson, Lars E. O. 1999. Price-level targeting versus inflation targeting: A free
lunch. Journal of Money, Credit and Banking 31:277–95.
Two Decades of Japanese Monetary Policy and the Deflation Problem
193
———. 2001. The zero bound in an open economy: A foolproof way of escaping
from a liquidity trap. Monetary and Economic Studies 19 (S-1): 277–312.
———. 2003. Escaping from a liquidity trap and deflation: The foolproof way and
others. NBER Working Paper no. 10195. Cambridge, MA: National Bureau of
Economic Research, December.
Svensson, Lars E. O., and Michael Woodford. 2003. Optimal policy with partial information in a forward-looking model: Certainty-equivalence redux. NBER
Working Paper no. 9430. Cambridge, MA: National Bureau of Economic Research, January.
Taylor, John B., ed. 1999. Monetary policy rules, 405–30. Chicago: University of
Chicago Press.
Taylor, John B. 2001. Low inflation, deflation, and policies for future price stability. Bank of Japan, Monetary and Economic Studies (special edition) February:
35–51.
Ueda, Kazuo. 1999. Remarks. Fed Conference, Monetary Policy in a Low-Inflation
Environment. 20 October, Woodstock, Vermont. Available at [www.boj.or.jp/en/
press/04/press_f.htm]
———. 2000. The transmission mechanism of monetary policy near zero interest
rates: The Japanese experience 1998–2000. Speech given at Japan Project Meeting, 22 September, Swedish Embassy in Tokyo. Available at [http://www.boj
.or.jp/en/press/04/press_f.htm]
Vestin, David. 2000. Price level targeting versus inflation targeting in a forward
looking model. Institute for International Economic Studies, Stockholm University, May. Mimeo.
Williams, J. C. 1999. Simple rules for monetary policy. Finance and economics discussion series no. 1999-12. Washington, DC: Board of Governors of the Federal
Reserve System, February.
Woodford, Michael. 1999. Optimal monetary policy inertia. NBER Working Paper no. 7261. Cambridge, MA: National Bureau of Economic Research, July.
———. 2000. Pitfalls of forward-looking monetary policy. American Economic Review 90 (2): 100–4.
———. 2003. Interest and prices: Foundations of a theory of monetary policy.
Princeton, NJ: Princeton University Press.
Yamaguchi, Yutaka. 1999. Monetary policy and structural policy: A Japanese perspective. Deputy Governor of the Bank of Japan, before Colloque Monétaire International at Banque de France, October 8–9. Paris. Available at [http://www
.boj.or.jp/en/press/04/press_f.htm].
Comment
Kenneth Kuttner
The broad objectives of this vast chapter are twofold. The first is to provide
a definitive overview of Japanese monetary policy since the mid-1980s. The
Japanese experience is then used to motivate its second objective, which is
to survey recent macroeconomic research on the deflation problem, and
using this, to propose policies to help solve Japan’s deflation problem. With
Kenneth Kuttner is the Danforth-Lewis Professor of Economics at Oberlin College, and a
faculty research fellow of the National Bureau of Economic Research.
194
Takatoshi Ito and Frederic S. Mishkin
two decades of economic history and an extensive literature to cover, the
chapter’s scope is a little overwhelming. The goals of these comments,
therefore, are to highlight and clarify some of its major themes, and to
identify a few areas where the chapter makes particularly valuable contributions. A number of unresolved issues deserving of further investigation
will also be noted.
The first half of the chapter, consisting of sections 4.2 through 4.4, summarizes the policies pursued by the BOJ over the past twenty years, and
some of the many critiques of those policies. Section 4.2 deals with the period prior to the BOJ’s formal independence, granted in 1998, while section
4.3 takes up the performance of the independent BOJ. Section 4.4 presents
and discusses the pitfalls of an effort to assess the BOJ’s policies using an
empirical policy-reaction function.
The detailed chronology of the BOJ’s policies contained in sections 4.2
and 4.3 stands as one of the chapter’s major accomplishments, and is likely
to serve as an important reference for years to come. While many of the
episodes covered in the Ito-Mishkin chronology have been discussed elsewhere, the presentation in this chapter is both comprehensive and compact, and benefits from Ito’s firsthand policy experience while serving in
the Ministry of Finance. Along with the narrative, sections 4.2 and 4.3 consider a number of the critiques that have been made of BOJ policies over
the years. These include the BOJ’s failure to act more aggressively to prevent the asset-price bubble of the late 1980s, its slow response to the onset
of the recession in the early 1990s, the reversal of the ZIRP in 2000, and the
Bank’s failure to “manage expectations” effectively.
Overall, the authors are relatively sympathetic to the BOJ’s predicament
in the late 1980s, acknowledging that the strong Yen and generally low inflation rates presented the bank with a dilemma that, to some extent,
explains its restrained response to asset-price inflation. Others, such as
Jinushi, Kuroki, and Miyao (2000), are more critical of BOJ policy during
this period, arguing that the bank’s misplaced emphasis on the exchange
rate kept it from responding to real economic conditions which, by the late
1980s, would have called for significantly tighter policy. Regardless of the
assessment, clearly this episode represents an interesting case study in the
potential pitfalls of exchange rate management, and probably deserves a
more detailed analysis.
The bulk of the authors’ criticism is reserved for the bank’s Hayami-era
policies—the premature abandonment of the ZIRP in 2000 is labeled a
“clear policy mistake,” and in the authors’ view, this (and other missteps)
have “left the Bank of Japan with a severe credibility problem in which the
markets and the public are unconvinced that Japanese monetary policy can
be committed to future expansion that would return the economy to
health.” The authors’ assessment of this period accurately reflects the consensus view, at least among academic economists, that monetary policy
Two Decades of Japanese Monetary Policy and the Deflation Problem
195
could have and should have done more to end the country’s economic stagnation. In emphasizing the potential role for monetary policy, however, the
chapter gives scant attention to supply-side explanations for the slow pace
of economic growth, such as that of Hayashi and Prescott (2002).
Section 4.4 supplements the narrative approach of sections 4.2 and 4.3
with a quantitative analysis based on empirical reaction-functions, or generalized “Taylor rules.” There are two basic ways reaction functions are
used to assess monetary policy. The first, referred to by Kuttner and Posen
(2004) as the “calibration” approach, is to insert output gap and inflation
measures into a reaction function with calibrated parameters (typically
Taylor’s) imposed a priori; if the implied interest rate path is lower than the
actual policy rate, policy is deemed “too tight.” The alternative, “estimation” approach is to estimate the reaction-function parameters and compare those estimates to “good” parameter values, such as Taylor’s; in Japan’s context, the interesting question is the extent to which policy did (or
did not) react to output fluctuations.
Unfortunately, reaction-function analysis of this sort is subject to a
number of methodological problems and pitfalls, as discussed in detail in
Kuttner and Posen (2004). Chief among these is the choice of output gap
measure, an issue that is especially germane to the case of Japan, where
most simple time-series methods (including those used by Ito and Mishkin) would spuriously attribute a significant portion of a prolonged cyclical slump to a reduction in trend growth. Not surprisingly, policy assessments consequently tend to be very sensitive to the choice of method used
to estimate potential output.
Ito and Mishkin run headlong into these problems in their efforts to assess BOJ policy. Their approach, which is to estimate a Taylor rule up
through 1994 and calculate projections from 1995 onward, yields results
that are plausible in some dimensions but odd in others. While the results
suggest the BOJ should have run a tighter policy in the late 1980s, policy
during the critical 1992–95 period is judged to be “about right”—and implies a significant tightening of policy in 1996–97. Calling these results
“counterintuitive,” the authors end up largely rejecting the Taylor-rule approach as uninformative, at least for the case of Japan. This conclusion is
surely well founded, for all the reasons listed in the chapter, as well as those
outlined in Kuttner and Posen (2004).
Nonetheless, the exercise contains one important result that deserves
additional emphasis: the lack of a significant response to the output gap in
the specification with a forward-looking inflation measure. A similar lack
of output response has also been reported by Kuttner and Posen (2001);
Jinushi, Kuroki, and Miyao (2000); and Ahearne et al. (2002). Estimates
for the U.S. Federal Reserve, on the other hand, such as those reported by
Clarida, Gali, and Gertler (2000), typically show an economically meaningful response to output gap fluctuations. The Fed’s sharp rate cuts in
196
Takatoshi Ito and Frederic S. Mishkin
2001–03 are consistent with such a response, as the 5.5 percentage point
rate reduction cannot be rationalized purely as a response to a drop in expected inflation. (With an inflation coefficient of 1.5, the observed rate reduction would imply an implausibly large 3.7 percent decline in inflation
expectations.) Thus, the lack of attention to real economic conditions revealed by the estimated reaction function may well have contributed to
Japan’s economic malaise.
The chapter’s second half (sections 4.5 through 4.7) is primarily a survey
of recent research on deflation and the ZLB problem. In its survey of the
literature, the chapter covers relatively well-trodden ground; it adds value,
however, in interweaving its survey with a discussion on how this research
might apply to the case of Japan. Drawing on this research, the authors recommend the adoption of a price-level target designed to bring prices back
up to where they would have been had there been a steady, low rate of inflation. This sensible proposal can be interpreted as a simplified version of
the rule advanced by Eggertsson and Woodford (2003), and it is rationalized by a similar set of considerations.
The BOJ has over the years strenuously resisted calls for an explicit price
or inflation target, however; and despite BOJ Governor Fukui’s apparently
more sympathetic attitude to reflationary policies, the likelihood of the
Ito–Mishkin proposal being adopted surely remains slim. In the past, the
BOJ officials have argued that, because the bank lacked the tools to achieve
an inflation target, the announcement of a target would undermine its
credibility. There is a certain undeniable logic to the view that announcing
a target is futile without the means to achieve it. It is worth stressing, however, that this pessimistic assessment requires either that “unconventional”
policies are ineffective (a view that the disappointing experience with quantitative easing has done little to dispel), or that the ZLB will always bind.
Searching for a deeper explanation, Ito and Mishkin offer three hypotheses for the BOJ’s resistance to an inflation target: first, BOJ officials’ belief
in the “cleansing” value of restrictive policy; second, a concern that a commitment to reflation would undermine its newly won independence; and
third, a fear that unconventional or overly expansionary policies would
jeopardize the bank’s balance sheet.
These are not the only possibilities, however. The authors might add to
their list of hypotheses the possibility that the BOJ may have been using
monetary policy as a weapon in its tussles with the Ministry of Finance
over fiscal policy. Alternatively, BOJ officials may actually have believed
that Japan’s deflation was “good,” a symptom of enhanced productivity
and efficiency. BOJ officials have also on occasion voiced concerns that
once inflation began, it would quickly accelerate, as a result of the large
“overhang” of liquidity in the financial system.
Some of these obstacles are probably impervious to economic reasoning; if policy is being handicapped by bureaucratic infighting, then there is
Two Decades of Japanese Monetary Policy and the Deflation Problem
197
not much that academic economists can do about it. But other obstacles,
such as the threat of insolvency or a liquidity overhang, merit a more detailed rejoinder. Perhaps in a sequel to this chapter, the authors will take up
some of these arguments and refute them in a more systematic fashion—
and in doing so, advance the cause of sound monetary policy in Japan.
References
Ahearne, Alan, Joseph Gagnon, Jane Haltimaier, and Steve Kamin. 2002. Preventing deflation: Lessons from Japan’s experience in the 1990s. Board of Governors of the Federal Reserve System, International Finance Discussion Paper
no. 729, June. Washington, DC.
Clarida, Richard, Jordi Galí, and Mark Gertler. 2000. Monetary policy rules and
macroeconomic stability: Evidence and some theory. Quarterly Journal of Economics 115 (February): 148–80.
Eggertsson, Gauti, and Michael Woodford. 2003. The zero bound on interest rates
and optimal monetary policy. Brookings Papers on Economic Activity Issue no. 1:
139–211.
Hayashi, Fumio, and Edward C. Prescott. 2002. The 1990s in Japan: A lost decade.
Review of Economic Dynamics 5 (1): 206–35.
Jinushi, Toshiki, Yoshihiro Kuroki, and Ryuzo Miyao. 2000. Monetary policy in
Japan since the late 1980s: Delayed policy actions and some explanations. In
Japan’s financial crisis and its parallels to U.S. experience, ed. Mikitani and Posen,
115–48. Washington, DC: Institute for International Economics.
Kuttner, Kenneth N., and Adam S. Posen. 2001. Inflation, monetary transparency,
and G3 exchange rate volatility. In Adapting to financial globalisation, ed.
Hochreiter and Hennessy, 229–59. London: Routledge.
———. 2004. The difficulty of discerning what’s too tight: Taylor rules and Japanese
monetary policy. North American Journal of Economics and Finance 15:53–74.
Comment
Kazuo Ueda
The chapter by Ito and Mishkin consists of four parts. First, it discusses the
macroeconomics of the so-called lost decade. Second, it provides a narrative description of the BOJ’s monetary policy during the last decade or two.
Third, it presents an analysis of the Taylor rule as applied to the Japanese
situation and shows the difficulties of using it as a guide for monetary policy. Fourth, it discusses the pros and cons of nonconventional operations
that may be used close to the ZLB on nominal interest rates.
My quick reaction to the authors’ analyses is: the discussion of the
macroeconomics of the lost decade misses some important points. The narrative account of the BOJ’s policy is partly accurate, partly unfair. The
Kazuo Ueda is one of the appointed members of the Monetary Policy Board of the Bank
of Japan, and a professor of economics at Tokyo University.
198
Takatoshi Ito and Frederic S. Mishkin
Fig. 4C.1
Rate of inflation in Japan
Source: Ministry of Public Management, Home Affairs, Posts and Telecommunications.
authors, by their focus on the Taylor rule, seem to have some misunderstanding about what the BOJ has been doing. Finally, the section on unconventional operations is well written, although I do not quite buy the authors’ conclusion, which seems to be just go ahead and do them no matter
what the costs are. I am afraid that, given my current position, I am unable
to comment in detail on the second part of the chapter. Hence, in the following I will discuss the remaining three parts of the chapter.
Let me start with the discussion of the macroeconomics of the Japanese
economy. The authors assert that macro deflation has been the number one
enemy of the economy. This seems to miss the point. Figure 4C.1 shows
Japan’s deflation of the CPI. The deflationary tendency did not set in until
the late 1990s when macroeconomic problems of the economy were already
apparent. In addition, the deflation has been mild. In figures 4C.2 and 4C.3
I show estimates of ex post real interest rates in Japan for the Great Depression period and for the post-1990 period. Unlike in the former, there is
no tendency for the real interest rate to rise with deflation in the late 1990s.
That is, no serious debt-deflation type dynamic was taking hold in the post1990 era.
Instead, the stagnation of the economy during the era seems to have
been due to the excesses—excess capital, labor, debt, and so on—built up
during the bubble period and negative financial accelerators generated by
the sharp fall in asset prices as argued in Baba et al. (2004). The authors
seem to realize the importance of these factors, but put too much weight
on the negative effects of macro deflation.
I now turn to the discussion of policy measures recently adopted by the
BOJ. The authors’ focus on the technical aspects of the Taylor rule seems
Two Decades of Japanese Monetary Policy and the Deflation Problem
Fig. 4C.2
199
Estimates of real interest rates, 1922–1935
Sources: Economic and Social Research Institute, Cabinet Office; Bank of Japan.
Fig. 4C.3
Estimates of real interest rates, 1991–2003
Note: Real interest rates are calculated as gross interest payments divided by total debt minus
the rate of increase in the deflator for domestic demand.
Source: Economic and Social Research Institute, Cabinet Office.
to indicate that they are of the view that the BOJ has been following the rule
in setting monetary policy. This is simply not correct. The Taylor rule is a
useful benchmark. The BOJ has not, however, blindly followed the rule. In
fact, as the authors correctly point out, it is possible to produce a fairly
wide range for the optimal level of the policy interest rate using the Taylor
rule, depending on assumptions about the parameters of the rule or about
the method to calculate the output gap.
Focusing too much on the Taylor rule is, however, problematic not so
much because of the difficulty of its implementation, but because the BOJ
200
Takatoshi Ito and Frederic S. Mishkin
has attempted to use measures that produce stronger easing effects than,
say, the Taylor rule. Under the Taylor rule, a central bank keeps the interest rate at zero so long as the rule says the rate should be either zero or negative. In order to produce stronger easing effects, the BOJ has made commitments regarding the duration of a zero rate. In April 1999 the BOJ
announced that it would “keep the overnight rate at zero until deflationary
concerns are over.” The framework was in place until August 2000. Under
current quantitative easing, ample liquidity provision—effectively a zero
rate—is promised to continue until actual and expected CPI inflation turn
positive.
The essence of such an approach is stated in Krugman (1998). Once the
ZLB is hit due to, say, a large exogenous decline in the natural rate of interest, a further increase in the money supply today has no effect on the economy. Assuming, however, that there is a nonzero probability that the economy is pushed out of the ZLB as a result of an exogenous rise in the natural
rate of interest tomorrow, a promise today of monetary expansion tomorrow will raise inflation expectations today and stimulate aggregate demand.
It is at once apparent that such a commitment ought to be stronger than
what the market naturally assumes about future monetary policy in the absence of the commitment. Otherwise, it will not affect expectations. There
are two ways to achieve this end. One is the announcement of a very high
inflation target. As Krugman puts it, “the central bank needs to announce
that it will be irresponsible.” The other is to commit to, in the event of a rise
in the natural rate of interest, slower increases in the interest rate than a
baseline monetary policy, say, the Taylor rule, suggests. In this case, the target rate of inflation does not have to be very high; however, the possibility
of inflation temporarily overshooting its target needs to be tolerated.
Clearly, what the BOJ has been doing is closer to the second of the two approaches. In any case, these approaches are already “nonconventional” in
the sense that they have not been employed elsewhere except by the BOJ
and, to a lesser extent, recently by the Fed. Baba et al. (2004) shows that
they have had some significant effects on the term-structure of interest
rates. It is also important, however, to recognize the limitations of the approach.
Essentially, the approach requires forces other than monetary policy for
stimulating the economy. As a result, it can become very strained when
such forces are weak. In the Krugman version, a lower probability of the
economy moving out of the ZLB tomorrow requires a correspondingly
higher inflation target. Very soon, the target becomes incredible because of
time inconsistency problems.1 The second approach, for example, raising
1. To put this differently, if the announcement of high inflation targets is credible, there is
no limit to the power of a permanent increase in base money to stimulate the economy. This
point is made in the McCallum chapter in this volume.
Two Decades of Japanese Monetary Policy and the Deflation Problem
201
interest rates more slowly than the Taylor rule as the economy improves,
only starts to exert large effects on medium- to long-term interest rates
when such improvements in the economy are expected to occur very soon.
A central bank may have to wait a desperately long time before such improvements take place. Improvements in the economy may not materialize
under the policy board that made the decision. As King (2004) points out,
the difficulty here is one of “collective decisions today may fail to bind future collective decisions.” In any case, these points have to be at the center
of discussion in any evaluation of the BOJ’s recent monetary policy.
Finally, Ito and Mishkin’s discussion of the pros and cons of other nonconventional monetary-policy measures in the neighborhood of the ZLB
is very reasonable. If I might add some obvious points, foreign exchange intervention may help in raising the price level. In the Japanese context, however, the BOJ is not allowed to carry out exchange rate policy, which is in
the hands of the MOF. The MOF, who has the power to carry out large interventions, ironically does not have the mandate to maintain price stability. Ito and Mishkin note various problems with operations in private real
assets. I agree with most of what they say here. I would have then thought
that the decision on whether or not to use such operations would be a function of how desperate the situation of the economy was. In fact, the BOJ
has been buying asset-backed securities since 2003 as a monetary-policy
action, and equities from private banks since 2002 as a prudential policy
measure. These reflected the BOJ’s judgment that the economy was in a serious situation, hence the use of some risky operations were justified, but
that the situation was not desperate, therefore, they should be carried out
with an eye to minimizing the negative effects of such operations on private
resource allocation.
References
Baba, N., N. Oda, S. Nishioka, M. Shirakawa, K. Ueda, and H. Ugai. 2004. Japan’s
deflation, problems in the financial system and monetary policy. Paper presented
at the 3rd Bank for International Settlements (BIS) annual conference, Brunnen,
Switzerland.
King, M. 2004. The institutions of monetary policy. NBER Working Paper no.
10400. Cambridge, MA: National Bureau of Economic Research, April.
Krugman, P. 1998. It’s baack: Japan’s slump and the return of the liquidity trap.
Brookings Papers on Economic Activity, Issue no. 2.